final report V8

Transcrição

final report V8

Barriers to competition in the supply of ECNS
1
Barriers to competition in the
supply of electronic
communications networks and
services
A final report to the European Commission
The opinions expressed in this Study are those of the authors and do not necessarily reflect the
views of the European Commission
©ECSC-EC-EAEC Brussels-Luxemburg 2003
David Lewin
David Rogerson
Assisted by:
Peter Alexiadis
Miranda Cole
November 2003
CLJ21
Final version
Ovum
November 2003
2
Executive summary.................................................................................................5
I
Introduction .................................................................................................5
II
Fixed network services ................................................................................5
III
The mobile network services markets ..........................................................7
IV
Broadcasting services .................................................................................9
V
e-services .................................................................................................11
VI
Software supply for ECNS.........................................................................12
VII
Policy issues: Separation of the incumbent................................................12
VIII
Policy issues: Measures for effective competition ......................................13
1
Introduction .............................................................................................16
1.1
The scope of the study ..............................................................................16
1.2
Our approach to the analysis.....................................................................18
1.3
The basis for our findings ..........................................................................19
2
Fixed network services ...........................................................................21
2.1
The supply chain for fixed network services ...............................................21
2.2
Competition in corporate markets ..............................................................23
2.3
Competition in narrowband fixed services..................................................24
2.4
Mass broadband services – the state of competition ..................................30
2.5
Mass broadband – infrastructure based competition ..................................34
2.6
Mass broadband – service based competition ...........................................40
2.7
Public WLANs ...........................................................................................53
2.8
Infrastructure vs service based competition ...............................................54
2.9
The move to next generation networks ......................................................55
3
Mobile network services .........................................................................59
3.1
Introduction ...............................................................................................59
3.2
Current industry structure and trends.........................................................59
3.3
The value chain for data services ..............................................................61
3.4
Major players and their market shares .......................................................62
3.5
Main market trends....................................................................................64
Ovum
November 2003
3
3.6
The impact of consolidation .......................................................................67
3.7
International roaming.................................................................................68
3.8
Spectrum trading.......................................................................................69
3.9
Use of SIMLOCK.......................................................................................69
3.10
MVNO hosting...........................................................................................70
3.11
Access for value added service providers ..................................................70
3.12
Supply conditions for greenfield 3G operators............................................71
3.13
Restrictions on network sharing.................................................................71
3.14
Roll-out conditions for 3G networks ...........................................................72
3.15
The use of walled gardens for data services ..............................................73
3.16
Dominance in the mobile terminal markets ................................................73
3.17
Cost based call termination prices .............................................................74
3.18
The impact of data protection on m-services..............................................75
3.19
Mobile content...........................................................................................76
4
Broadcasting services ............................................................................78
4.1
The broadcasting services value chain ......................................................79
4.2
Potential Barriers to Competition From Rights Acquisition and Exercise.....83
4.3
Consistent Regulatory Treatment of Competing Services ..........................89
4.4
Access to Conditional Access Systems .....................................................92
4.5
Interoperable iTV services and APIs..........................................................94
5
e-services ..............................................................................................107
5.1
The current industry structure..................................................................107
5.2
Corporate e-services ...............................................................................108
5.3
Backbone ISPs........................................................................................109
5.4
Consumer oriented ISPs .........................................................................111
5.5
e-payment issues ....................................................................................113
6
Software supply for ECNS ....................................................................131
6.1
Introduction .............................................................................................131
6.2
Wireless handset operating systems........................................................131
6.3
Wireless middleware ...............................................................................133
6.4
Web services ..........................................................................................134
Ovum
November 2003
4
6.5
Browsers for mobile devices....................................................................135
7
Separation of the incumbent ................................................................139
7.1
Options for separation .............................................................................139
7.2
Divestiture of Loopco...............................................................................140
7.3
Divestiture of Netco .................................................................................140
7.4
Divestiture of CATV businesses...............................................................141
7.5
Divestiture of mobile subsidiaries.............................................................146
8
Measures for effective competition......................................................150
8.1
Introduction .............................................................................................150
8.2
Definitions ...............................................................................................150
8.3
Infrastructure-based competition is better ................................................151
8.4
Limitations on infrastructure-based competition .......................................153
8.5
Prospects for infrastructure-based competition ........................................154
8.6
The role of service-based competition .....................................................156
8.7
Maximising infrastructure-based competition ...........................................157
8.8
Increasing infrastructure investment ........................................................158
8.9
Stimulating cross platform competition.....................................................158
8.10
Increasing investment incentives for Altnets.............................................160
Annex A Directives affecting micropayment services.....................................166
The E-money Directive..........................................................................................166
The Electronic Signatures Directive.......................................................................168
The Codified Banking Directive .............................................................................169
The Money Laundering Directive ...........................................................................169
Annex B Glossary .............................................................................................171
Ovum
November 2003
5
Executive summary
I
Introduction
This study prepared by Ovum1 looks at barriers to competition in the supply of
electronic communications networks and services (ECNS) in the EU. The opinions
expressed in this Study are those of the authors and do not necessarily reflect the
views of the European Commission.
The main market areas covered are fixed network services, mobile services, television
services, and Internet based services (e-services). The study also looks at the supply of
software for 3G mobile services.
II
Fixed network services
No barriers to competition in the corporate market
Rivals to the fixed incumbent in the corporate market have won very substantial shares in
providing services based on alternative infrastructure in city centres. Compared with other
segments of the fixed network services markets there are relatively few barriers to
competition.
We recommend that the European Commission should take no further action in relation to
the corporate sector until it is clear that current measures are failing.
Competition in narrowband fixed services
Fixed incumbents continue to dominate in the supply of narrowband access and local calls
in most member states. This reflects the very substantial barriers to competition which
2
exist in this market . In contrast incumbents face strong and growing service based
competition in the long distance calls market.
The lack of competition in the narrowband access market has led several NRAs to mandate
that the fixed incumbent supply wholesale line rental (WLR) services. For WLR the
incumbent rents the narrowband access lines and associated features at the local switch to
a rival at a regulated wholesale price. Our analysis indicates that the balance of arguments
in favour of mandating WLR service is far from overwhelming. One major objection is that it
removes incentives for infrastructure based competition.
1
A team lead by David Lewin and David Rogerson from Ovum and Peter Alexiadis
and Miranda Cole, then of Squire, Sanders & Dempsey LLP and now of Gibson,
Dunn & Crutcher LLP.
2
In terms of the economies of scope and scale, sunk cost investment and
advantages of incumbency enjoyed by the old monopoly operator.
Ovum
November 2003
6
We recommend that Member States investigate further whether WLR is effective in
removing barriers to competition or whether it acts to deter infrastructure competition.
Infrastructure based competition in broadband services
Demand for mass broadband service is growing rapidly. By the end of 2007 we expect 90
million of the 200 million fixed connections in the current EU to operate at broadband
speeds3. The level of competition in this sector varies considerably between member
states and the EU lags far behind trading rivals like the USA, Japan or Korea in the level of
infrastructure based competition4.
Given the financial difficulties which major European CATV operators have faced over the
past few years we do not expect many of them to expand their geographic footprint
substantially from their current coverage. But these operators still offer a powerful source of
infrastructure based competition to the fixed incumbent in the broadband markets. A major
barrier to competition is the fact that, in five member states, the fixed incumbent still owns a
substantial proportion of CATV network facilities.
We recommend that the EU and the Member States consider the steps that they can take
to ensure that fixed incumbents divest themselves of their CATV businesses
At the same time the prospects of infrastructure based competition from new technology
networks are mixed. Our research suggests that:
• prospects for competition based on fixed wireless access is very limited; but that
• there are new prospects for powerline communications (PLC) in which
telecommunication signals are carried over the electricity distribution network.
Given PLC’s potential as a source of infrastructure based competition we believe it is
important to resolve the interference problems which PLC equipment currently faces.
State funding of broadband rollout could also affect infrastructure based competition.
Virtually all respondents to our study are convinced that guidelines on the application of
state funding are being ignored and that this is leading to a distortion of competition.
We recommend that the European Commission should investigate these complaints and
take appropriate action.
Service based competition in mass broadband services
Incumbents currently offer three broad types of wholesale products - IP level wholesale
DSL service, bitstream DSL service, and unbundled local loops.
We recommend that:
• SMP operators should be required to offer all three services.
3
Speeds in excess of 128kbit/s.
4
If we count unbundled local loop operators as infrastructure competitors then less
then 50% of broadband services are supplied by incumbents in these countries. In
the EU the proportion is around 75%.
Ovum
November 2003
7
• SMP operators should be required to offer practical and effective migration paths
between the products. This allows service providers to use the three products to build
their customer base, reduce risks and so provide “stepping stones” from service based
to infrastructure based competition.
• SMP operators should be required to provide co-mingling of equipment and costoriented backhaul at an appropriate level of disaggregation to local loop unbundlers.
Such measures will help lower the main barrier to local loop unbundling – the large
upfront investment required to reach the incumbent’s local loops – and so maintain the
rapidly growing proportion of broadband services provided using unbundled local
loops5.
• Where national regulatory authorities (NRAs) take decisions about prices for these
services, incentives for the incumbent and its rivals to rollout broadband infrastructure
need to be preserved. If a cost oriented price for a bitstream service is imposed, NRAs
will need to take account of investments in other new products which have failed, the
asymmetric risks which incumbents face when making sunk investments in new
technology products in competitive market places, and the rate of price/performance
improvements of new technology components.
Detailed analysis can be found in the main report.
III
The mobile network services markets
Industry structure
The mobile industry structure is very different from that in the fixed services industry. While
the fixed incumbents still provides well over 95% of fixed connections in most Member
States, the largest mobile operator typically has a market share of between 40% and 60%.
There are three to five 2G operators offering service in all member states with an additional
greenfield 3G licensee in many. In 13 of the 15 members states the leading mobile operator
(MNO) is a subsidiary of the fixed incumbent. There are in addition a number of
independent service providers reselling MNO services and a growing number of
aggregators, content providers and applications developers involved in the supply chain for
data services.
There is a general, but not universal, support for the thesis that we will see consolidation
in the EU mobile industry over the next few years as a result of economy of scale effects.
However, there is also a general view that the level of infrastructure competition in the
supply of mobile network services will remain satisfactory in most EU Member States after
consolidation and that competition law provides a satisfactory mechanism for dealing with
any competition issues which may arise from consolidation in these Member States.
We conclude that:
5
In October 2002, this proportion was at 2.5%. By June 2003, it had reached 4.7%.
Ovum
November 2003
8
• there is no need for the European Commission to institute ex ante EU-wide measures
to deal with competition problems arising from consolidation of the mobile industry; and
• it is important that NRAs do not intervene to preserve non-viable mobile operators who
come to them seeking regulatory relief.
Access for MVNOs and value added service providers
There are a growing number of calls from independent service providers, MVNOs and
value added service providers to gain access to the customers and services of the MNOs
on regulated terms. But there are also arguments against introducing such regulation. In
particular:
• it is important to preserve the current level of infrastructure based competition
within the mobile industry, especially given the prospects for consolidation
discussed above. Mandatory access for service providers would reduce future
investment incentives for mobile operators;
• there are strong market incentives for mobile operators to reach satisfactory
commercial arrangements with service providers, especially now that spare capacity on
3G networks is becoming available;
• it is clear that many mobile operators, especially the smaller operators, are following an
open approach to data and Internet services, rather than a walled garden approach;
and
• a walled garden approach can have value in helping to build user confidence in new
data services and in offering consumer protection. Mobile operators can make sure that
walled gardens services are easy to use, offer predictable prices and minimise the risk
of fraud6.
3G Roll Out
There are serious problems of site acquisitions and site sharing, especially for greenfield
3G operators in 3G roll out. However the problems are complex and vary across member
states.
We can see little merit in preserving restrictions on network sharing. So we recommend
that the European Commission should examine whether it is practical for the restrictions on
network sharing which exist in many Member States to be removed.
Should 3G roll out conditions be relaxed? According to mobile operators roll out conditions
are not especially onerous, except in Sweden. Even here it should be possible, given that
W-CDMA cell sizes depend on levels of use, to interpret licence conditions flexibly so as to
align them with current projections of market demand.
Assuming that such a flexible approach is possible, we conclude that there is no case for
EU-wide action to modify existing 3G roll-out conditions.
6
A mobile operator which offers only a walled garden approach could reduce the
level of competition and will reduce consumer choice.
Ovum
November 2003
9
Regulating the prices of mobile call termination
Should mobile operators be subject to requirements for cost orientation and nondiscrimination when charging for their call termination services? The issue of cost
orientation is currently the subject of administrative and judicial review in a number of
member states. Accordingly, we do not believe that it is appropriate to take a position in
relation to this issue.
However, we do recommend that the European Commission should:
• consider whether it should issue guidance requiring NRAs to ensure that mobile
operators charge regulated call termination prices in a non-discriminatory fashion and,
in particular, do not discriminate between charges to rivals and charges to their own
retail business.
• consider what price differences between mobile operators should be allowed when
setting a regulated call termination price. For example, should differences in call
volumes, which lead to higher unit costs for smaller operators, be reflected in the price
differences?
• consider to what extent the arguments which lead to regulated prices for 2G mobile
voice termination should apply to other services such as SMS call termination, 3G
voice call termination and MMS call termination.
IV
Broadcasting services
Potential barriers to competition from rights acquisition and exercise
Control of rights to premium content may have a foreclosing effect, particularly where such
rights relating to a wide variety of content are held by individual platform operators. The
same effect may occur in relation to the licensing of rights to TV programme content with
embedded enhanced functionality.
We therefore recommend that the European Commission should:
• use competition law and its attendant procedures (where appropriate) to limit exclusive
rights to TV programme content, particularly where such rights are of broad scope
(both in terms of platforms and number and variety of rights) and are of long duration.
• consider developing guidelines relating to content rights, based on case law to date.
• study further the issues associated with embedded enhanced content to see whether
broadcast content and embedded enhanced content should be considered as separate
markets.
Relationship between channel supply and access to platforms
Such relationships are the interface between content and the regulatory regime applying to
electronic communication services (ECS). Within the regulatory framework, there is
variation in the treatment of substitutability between transmission on cable and satellite
platforms. There appears to be more consistency in the treatment of analogue terrestrial
Ovum
November 2003
10
broadcasting transmission services as complements to these other transmission services.
It is not clear how the remedies that might be imposed on a provider of broadcasting
transmission services with SMP under the ECNS regulatory framework will operate in
practice. For example, it is not clear whether an “access” obligation would effectively
confer on content providers a right to access to transmission services, even though they
are not otherwise within the scope of the ECNS regime.
We recommend that the Commission and the Member States consider carefully the
variation between Member States when identifying relevant broadcasting transmission
markets and consider the appropriate form of obligations to be imposed on entities found to
have SMP in the relevant market(s).
Consistent Regulatory Treatment of Competing Services
The nature and scope of must carry obligations currently vary widely across Member
States. The regulatory framework is unlikely to lead to a review of channel bouquet lineups and pricing decisions currently applied in a number of Member States.
The Commission should further investigate the related legal issues to ensure that ‘must
carry’obligations remain appropriate and proportionate.
The regulatory framework does not apply to services providing or exercising editorial
control over content transmitted over the networks. The Television Without Frontiers
regime uses the ‘television broadcasting’concept.7 The gaps and overlaps between these
concepts and the different impact of each regime is already beginning to open up potential
regulatory lacunae. These differences can be seen in relation to the scope of each regime,
the underlying jurisdictional basis and the powers conferred on regulators in relation to the
different services.
The Commission should undertake a comprehensive review of the overlaps of, gaps
between, and boundaries of, these regulatory distinctions to ensure that the application of
different regulation to different services does not defeat the Community's digital economy
aims.
Conditional Access Systems
Limited regulatory attention has been focussed on interpreting the “fair, reasonable and
non-discriminatory” access obligation under the regulatory framework. Any review of the
remedy should entail consideration of:
• whether an obligation on conditional access systems might provide the most
appropriate means to address the regulatory mischief;
7
Defined as the initial transmission of television programmes intended for reception
by the public, including communication of programmes between undertakings with a
view to their being relayed to the public. It does not include communication services
providing items of information or other messages on individual demand such as
telecopying, electronic data banks and other similar activities.
Ovum
November 2003
11
• the extent to which other aspects of service provision (e.g., access to transmission
services and/ or platforms and the impact of ‘must carry’obligations) impact on the
appropriateness and proportionality of remedies; and
• whether only ‘broadcasters’should be the beneficiaries of any access obligations.
NRAs and the Commission should consider carefully the likely impact of lifting the existing
fair, reasonable and non-discriminatory access obligations from providers without SMP and
the appropriateness and proportionality of the current and other potential access
obligations.
Interoperable iTV services
The regulatory framework legislation contemplates the possibility of mandating standards
for middleware to ensure the interoperability of iTV services (e.g., the ability of interactive
content to be ported between platforms with the minimum amount of reauthoring). We
have found that authoring tools facilitate the porting of content. However the following
factors reduce the portability of content and applications: transmission bandwidth
(particularly the return channel), network integration, processing resources (in the set-topbox), different APIs, transaction processing and linguistic and cultural issues.
Platform-independent content interchange formats provide enhanced portability through a
mechanism for describing applications in a platform-neutral manner (also called “authoring
at a high level”). Such an approach separates content and templates, allowing the
generation of platform-specific versions on services. Industry estimates that approximately
80% of current iTV applications could be authored using platform-independent languages.
The Commission should identify, before July 2004, the relevant market failures, with a view
to focussing on whether they relate to standardisation or interoperability. It should also
explore the extent to which authoring tools could address such failures.
V
e-services
Competition between e-service suppliers
There are no barriers to competition in the supply of global Internet connectivity by
backbone Internet Service Providers (ISPs) or in the supply of corporate e-services which
warrant regulatory intervention. However, there are problems in the supply of consumer
oriented e-services. Independent consumer oriented ISPs are seeking lower, cost oriented,
prices for bitstream access and complain about discrimination by fixed incumbents. These
ISPs claim that they will exit the market if there is no regulatory intervention. If true, this
leads to a difficult choice for NRAs. If they attempt to preserve service competition from
such ISPs by setting cost-based prices on a low level, they could significantly reduce
further infrastructure investment.
The European Commission should investigate claims of discrimination and take appropriate
action. The issue of access pricing is dealt with in Section I of this executive summary.
Ovum
November 2003
12
The role of e-payment mechanisms and regulation
The primary barrier to the introduction of e-payment services arises from the potential for
pre-paid micropayment systems to fall under the onerous and expensive regulations
applicable to e-money. The key factor that is accepted in some, but not all, Member States
as distinguishing such micropayment mechanisms and e-money is the requirement that the
latter be accepted by undertakings other than the issuer of the value. There is currently
wide variation in the interpretation of when value is ‘accepted’by a body other than the
issuer.
The European Commission and national authorities should consider issuing guidance as to
the scope of the application of e-money regulation to pre-paid micropayment systems. In
the meantime the relevant authorities should also consider guidance relating to:
• the appropriate and proportionate mechanisms to separate pre-paid value from issuer
funds
• the need for redeemability requirements in relation to small value pre-payments
• the proportionality and appropriateness of the money laundering requirements.
VI
Software supply for ECNS
Software suppliers will play an increasingly important part in the value chain of the ECNS
industry and in particular in the development of 3G mobile services through the supply of:
• operating systems and web browsers for mobile terminals;
• wireless middleware for mobile network services and, more generally;
• web services to enable different software systems to interoperate.
There are currently neither competition problems nor requirements for regulatory
intervention in these areas.
VII
Policy issues: Separation of the incumbent
Market power within the telecommunications service industry in the EU remains firmly
concentrated in the hands of those organisations which formerly provided service on a
monopoly basis. However, the incumbent’s various businesses could be established as
legally separate entities (legal separation) or as separately owned businesses (divestiture).
We have analysed where the costs of separation may be exceeded by the benefits it
brings. We find that:
• the case for divesting the incumbent’s fixed access network from the rest of its fixed
network business and running it as a separate entity is weak.
• the case for divesting the fixed network business of the incumbent from its retail
business is stronger, but still weak.
• there is a strong case for incumbents to divest themselves of their CATV operations .
• there is, as yet, no case to require incumbents to divest themselves of their mobile
subsidiaries.
Ovum
November 2003
13
VIII Policy issues: Measures for effective competition
In trying to enable the growth in public welfare benefits which competition in the ECNS
markets can bring, NRAs are constantly faced with the problem of trying to determine the
proper relationship between measures designed to promote infrastructure-based
competition and measures designed to promote service-based competition:
• where it is viable, infrastructure-based competition is better than service based
competition. It requires less regulatory intervention and so reduces the scope for
regulatory error and the economic costs which such errors generate. It also generates
more competitive pressure. Pure service-based competition puts pressure on the
incumbent in terms of retail efficiency, customer service innovation and price levels.
Infrastructure competition does the same, but creates additional pressures on the
incumbent to innovate in network services, to differentiate in terms of products and
pricing structures and to improve overall cost efficiency and quality of service. In other
words, infrastructure based competition generates dynamic benefits.
• unfortunately, infrastructure based competition is not always viable – especially in the
supply of fixed network services where there are substantial economies of scale and
scope and where the scale of the sunk investment required makes investment risks for
new entrants too high.
• in these circumstances, it makes sense for NRAs to introduce ex ante measures8 which
enable service based competition at the retail level. Measures designed to promote
service-based competition tend to have a rapid and visible effect on retail competition
in telecommunications. When priced appropriately, they can also provide a stepping
stone to infrastructure-based competition. By competing at the service level, entrants
can build a customer base and revenues with little investment risk and then migrate the
customers to their own facilities.
• however, where wholesale prices are set too low measures to promote service based
competition can have unintended and damaging consequences. For example, they can
undermine infrastructure based competition or, more importantly, they can undermine
infrastructure investment in new technology services.
Based on this analysis we recommend that NRAs should take account of the
following factors in determining the right balance between measures designed to
promote infrastructure and service based competition:
• the relatively poor prospects for investment in further infrastructure-based
competition in fixed services.
• the fact that a significant proportion of this competition is currently based on price
averaging by the incumbent, both in terms of geography and customer groups.
• the need to set regulated prices for new technology services which do not remove
the incentives for infrastructure investment by the incumbent and new entrants.
8
Such as mandatory carrier pre-selection and call origination services; requiring
wholesale line rental offerings or; supplying DSL access services at regulated prices
Ovum
November 2003
14
• the need to pursue a consistent long term regulatory policy towards infrastructurebased operators.
Given its superiority over service based competition we recommend for implementation 11
measures which are designed to maximise infrastructure based competition without
encouraging inefficient investment.
We propose two measures to increase infrastructure investment:
• regulatory authorities should ensure that they pursue a consistent long term
regulatory policy towards infrastructure-based operators.
• NRAs should set the prices of inputs which promote service-based competition at
levels which preserve incentives for infrastructure investment by incumbents and
their infrastructure based rivals, as discussed in Section I.
We propose four measures to increase cross platform competition:
• the EU and individual Member States should take what steps they can to require
fixed incumbents to divest themselves of any remaining interests in CATV
network operators.
• NRAs and national competition authorities in Member States where the fixed
incumbent owns the leading mobile operator, should monitor the possibility of
leverage between these two parts of the incumbent’s business closely. Where
anti-competitive conduct occurs, they should consider requiring structural
separation or divestiture.
• NRAs should ensure that the call termination charges of fixed and mobile
operators are regulated on a consistent basis .
• NRAs or national competition authorities should investigate whether mobile
operators are acting in an anti-competitive manner in the pricing of the services
which they offer to large corporate customers for the termination of their voice
traffic on public networks.
Finally we propose five measures to increase cross platform competition:
• NRAs should allow the infrastructure-based rivals to the incumbent – whether
fixed or mobile – to set call termination charges which allow them to recover the
efficiently incurred costs of an operator of their size and topology.
• NRAs should make an explicit, but not necessarily quantitative, assessment of
the costs and benefits of any regulatory measure which is designed to enable
service-based competition.
• NRAs should make their current policy on geographical averaging of the fixed
incumbent’s retail prices explicit. Then, before allowing the incumbent to
geographically de-average prices further, they should consider the likely impact
on infrastructure-based competition.
• given the relatively low level of infrastructure-based competition in fixed network
services in many member states, each NRA should consider whether it should
provide explicit entry assistance to infrastructure-based rivals to the fixed
incumbent.
Ovum
November 2003
15
• the relevant authorities should ensure that the EU rules on the use of state aid to
fund telecommunications investments are implemented rigorously.
Ovum
November 2003
1
16
Introduction
1.1 The scope of the study
In February 2003 the European Commission asked Ovum9 to carry out a wide
ranging review of barriers to competition in the supply of electronic communications
networks and services (“ECNS”) in the EU.
We have structured our analysis, and our report, to look at the five broad market
areas shown in the value chain diagram of Figure 1.1.
As Figure 1.1 shows our study ranges beyond the boundaries of ECNS, into
broadcast and Information Society Services and software markets. We consider it
important to explore these issues, so as to ensure that we identify barriers to
competition in ECNS markets which arise through leverage from related markets.
Figure 1.1 The scope of the study
CPE
supply
End users
Chapter 6
Fixed
network
services
(Chapter 2)
Mobile
network
services
(Chapter 3)
TV
broadcast
Chapters 4
Network
software
Chapter 7- divestiture by fixed incumbent
e-services (Chapter 5)
Chapter 8 - Infrastructure v service based
competition
Content supply
The individual chapters cover the following issues:
• in Chapter 2, we look at barriers to competition in the supply of fixed
network services. We divide our analysis into six main parts:
• barriers to competition in the supply of services to the corporate market
(Section 2.2).
9
A team lead by David Lewin and David Rogerson from Ovum and Peter Alexiadis
and Miranda Cole then of Squire Sanders and now of Gibson, Dunn and Crutcher
Ovum
November 2003
17
• barriers to competition in the supply of mass market narrow band services
(Section 2.3).
• barriers to infrastructure-based competition in the supply of mass
broadband access (Section 2.5).
• barriers to service-based competition in the supply of mass broadband
access (Section 2.6).
• competition problems associated with the supply of public WLANs (Section
2.7).
• competition problems associated with the move to next generation
networks (Section 2.9).
We also mention briefly the problem of establishing a proper relationship between
the service-based and infrastructure-based competition (Section 2.8). This
problem is analysed in detail in Chapter 8.
• in Chapter 3 we examine barriers to competition in the supply of mobile network
services. In particular, we consider the growing importance of data services, the
impact this has on supply chains and the potential barriers to competition which
then arise. We also discuss the possibility of consolidation in the EU mobile
industry as a result of economies of scale in the provision of service and the
regulatory issues which this consolidation raises.
• Chapter 4 considers barriers to competition in the supply of broadcast services.
We consider issues of exclusive content and the balance between content and
platform providers. We also raise for discussion the way in which the new
regulatory framework defines content, deals with must carry obligations and treats
access to conditional access systems. Finally we look in detail at the best way to
ensure interoperability for interactive TV services over application programme
interfaces
• Chapter 5 looks at potential barriers to competition in the various segments of the
e-services market. It covers consumer oriented ISPs, backbone ISPs, corporate
ISPs and e-payment service providers. It then looks in detail at the regulation of
micropayment services and the need to strike a balance between making these
services viable and protecting consumers from their misuse
• Chapter 6 analyses key software developments which will impact on
competition in ECNS markets over the next few years. We look at four main
areas:
-
the development of operating systems for mobile terminals
the emergence of wireless middleware
the evolution of Web Services
the markets for web browsers for PCs and mobile terminals.
In the last two chapters of our report we look at two competition problems which affect
the ECNS markets more generally:
• in Chapter 7 we consider the concentration of market power which currently lies
in the hands of the fixed incumbents of most member states and examine the
case for requiring the incumbent to divest itself of various businesses
Ovum
November 2003
18
• In Chapter 8 we look in detail at the problem of establishing the proper
relationship between infrastructure-based and service-based competition so
as to maximise effective competition in the telecommunications industry (fixed
and mobile) as a whole.
1.2 Our approach to the analysis
In each of Chapters 2 to 5 we have adopted the same general approach:
• to discuss the value chain of supply. See in particular Sections 2.1, 3.2, 3.3, 4.1
and 5.1.
• to examine the current state of competition in the supply of services. See in
particular Sections 2.3, 2.4, 3.4, and 3.5.
• to identify barriers to competition in this supply. These are distributed throughout
the text and our recommendations on how to deal with them are highlighted
through the use of boxes and italic text. We identify and analyse the main
barriers to competition in Sections 2.2 to 2.7 (fixed markets), Sections 3.6 to 3.18
(mobile markets), Sections 4.2 to 4.5 (broadcast markets) and Sections 5.2 to 5.5
(e-services markets).
• to consider and, where appropriate, to propose regulatory measures to remove
these barriers to competition. All the main measures are set out in, or cross
referenced from, Chapter 8.
We start our enquiry by looking for barriers to competition in the supply of services to
end-users. Where problems arise, we then look further upstream along the value
chain. We focus on:
• identifying new barriers to competition which are created as result of market
developments. This means that there is a focus on barriers to competition in the
provision of broadband services. These barriers to competition are covered
specifically in Sections 2.5 and 2.6 for the fixed network markets ;
• barriers to competition which arise because of consolidation within the industry
as players exit or merge with others. We discuss the consolidation of the mobile
industry in Section 3.6;
• barriers to competition which arise because regulatory initiatives have failed or
appear to be failing. Accordingly, we have excluded from our enquiry barriers to
competition for which there appear to be successful regulatory remedies such as
requirements to supply private partial circuits to enable greater competition in the
corporate sector;
• barriers to competition which will become significant over the next three years,
excluding barriers to competition which it appears might become important
beyond 2006. We include a discussion of the competition problems which might
arise from the migration to next generation IP networks in Section 2.9. But
otherwise we focus on relatively short term issues;
• barriers to competition in the provision of transport services. As such, we largely
exclude consideration of barriers to competition in the supply of content –
Ovum
November 2003
19
although Section 4.2 considers the important relationship between content
providers and platform operators.
The focus of our report is on analysing markets where there are clear competition
problems. This means that we pay relatively little attention to markets which are
functioning well or to new and fast-changing markets where competition problems
have yet to emerge. Many Internet based markets fall into these categories as our
analysis in Chapter 5 shows.
1.3 The basis for our findings
Our report is based on three main pieces of work :
• first we analysed the industry value chains and market developments for the main
ECNS market segments – fixed, mobile, TV broadcast and e-services. Based on
this analysis, we identified a wide range of possible barriers to competition.
• secondly we consulted to gather reactions from interested parties. We presented
our analysis in a discussion document which we circulated widely for comment,
and held meetings with key parties.
• thirdly we investigated certain topics – broadband wholesale pricing,
micropayment regulations, APIs, long term retail price control and divestiture of
the fixed incumbent - in more detail, carrying out 16 additional interviews as we
did so.
Overall we carried out over 40 interviews involving over 30 different organisations
which included user groups, NRAs, fixed incumbents, CATV operators, other
alternative access network operators, ISPs, mobile operators, TV broadcasters and
software suppliers. Figure 1.2 lists the organisations which provided inputs to the
study.
Ovum
November 2003
20
Figure 1.2 The organisations who contributed to our study
NRAs and NRA groups
IRG
Oftel
OPTA
Fixed incumbents
TDC
Telecom Italia
BT
Deutsche Telekom
Kingston Communications
Mobile operators
Hutchison 3G
Orange
Vodafone
O2
T-Mobile
GSM Association
ISPs
Tiscali
AOL
CATV operators
Telenet
UPC
Dutch CATV Association
ntl
Other Alternative network
operators
e.biscom
COLT
AT&T
Cable and Wireless
Software suppliers
Microsoft
Symbian
User groups
INTUG
TV service providers
Open TV
DVB
AOL Time Warner
Turner Broadcasting
BSkyB
UPC (CATV operator)
ntl (CATV operator)
Others
OECD
ECTA
EP Star – Finland
CEC experts on remedies
Ovum
November 2003
2
21
Fixed network services
2.1 The supply chain for fixed network services
The supply chain for the delivery of fixed network services to end-users is complex.
Following full liberalisation of the EU telecommunications markets in 1998, there are
now many thousands of service providers and network operators delivering fixed
network services to end-users and other service providers across the EU. Figure 2.1
provides a simplified version of these supply chains which attempts to capture the
essential features of both narrowband and broadband supply.
Figure 2.1 The supply chain for fixed network services
End users
Incumbent’s access
network
Indicates direction
of delivery of
a service
LLU
Incumbent’s
core network
facilities
ISPs
Rival’s
access
network
Access
Rival’s core
network facilities
Network services - incumbent and rivals
Resale
Resale
Narrowband switchless resellers
The main characteristics of this supply chain are as follows:
• economies of scale in the provision of access networks mean that the fixed
incumbent supplies the line connecting most customers to their
telecommunications network supplier;
• the nature of investment in access networks reinforces this trend. Alternative
access network operators must make substantial up-front investment before they
can start to build a customer base. Once made this investment is sunk – it cannot
be re-used for other purposes. This substantially raises the risk of market entry
when compared with the risks faced by incumbents – which in many cases had
already made a large proportion of their infrastructure investments while they
enjoyed a monopoly
Ovum
November 2003
22
• incumbents and their rivals use logically separate networks to deliver broadband
and narrowband services to end users. But these networks share use of access
network components, transmission ducts and transmission capacity to a
considerable extent
• the main alternative fixed access networks are supplied by companies which
specialise in serving the corporate market (e.g., Equant and COLT); by CATV
companies (e.g., Telenet, UPC and NTL) which serve the mass markets; and by
local access network providers such as NetCologne. There are also broadband
satellite operators providing alternative infrastructure;
• many alternative access network providers have built their own infrastructure.
But some rent unbundled local loops form the fixed incumbent, using either the
loop as a whole or, in some cases, the high frequency part of the loop10 to provide
broadband access
• incumbents dominate in the supply of narrowband access services11. But, as
shown in Figure 2.2, they now face substantial competition in the long distance
and international calls markets. They also face significant competition in the local
calls market in some Member States12. Regulatory initiatives which require
incumbent operators to offer cost-based call origination and carrier selection (and
preselection) to rivals have had considerable success here. These services entail
rivals purchasing core network services13 from the fixed incumbent;
• infrastructure competition in the supply of broadband access is stronger than in
narrowband access. But it is very variable across the EU. In several member
states well over 80% of broadband access services are supplied by the
incumbents; in several others the fixed incumbent supplies less than 50%. This
variation largely reflects the extent to which the fixed incumbent faces competition
from CATV operators
• ISPs, like Tiscali and AOL, buy access services at wholesale prices, largely from
fixed incumbents; and
• switchless resellers buy end-to-end services at wholesale prices, again, largely
from the fixed incumbent.
10
Shared access
11
With incumbents supplying over 95% of narrowband lines in most member states
12
Such as Italy.
13
Such as call origination
Ovum
November 2003
23
Figure 2.2 The market share of incumbents in key narrowband markets
Country
Market share (outgoing minutes)
Belgium
Denmark
local
long distance
international
Years since full
liberalisation
na
na
52%
6
na
na
53%
8
Finland
93%
na
54%
8
France
85%
62%
63%
6
Germany
96%
63%
48%
6
Greece
100%
99%
97%
3
Ireland
88%
59%
66%
5
Luxembourg
89%
na
72%
3
Netherlands
83%
76%
62%
5
Portugal
Spain
UK
Source: 8
TH
na
90%
70%
3
77%
84%
82%
4
76%
48%
30%
11
Implementation Report; information not available for Austria, Italy or Sweden
2.2 Competition in corporate markets
Competition in the provision of fixed network services – both narrowband and
broadband – to large corporate customers is much stronger than it is in the mass
market. It is much easier for service providers to justify direct connection to large
corporate sites14 given the substantial revenues that these sites generate. So
alternative network operators have built out extensive fibre networks in many city
centres and the incumbents in many Member States report market shares
significantly below 50% in central business districts. For example, BT, which has
faced competition longer than most incumbents in Europe, reports a market share of
less than 20% in the City of London.
There are problems, however, in that:
• operators which specialise in providing services to corporate customers over their
own facilities, such as Equant and Colt, generate EBITDA margins in the 5% to
10% range, whilst their incumbent rivals generate margins of approximately 30%.
They specify the high cost of connectivity to customer sites as the key reason for
their low profitability. There are few incentives for additional investment at such
margins and there is little sign of improved margins in the near term in the
projections of the financial analyst who follows these companies.
• companies which serve the corporate sector have also indicated that it is difficult
for them to compete for the business of large multi-site customers, particularly
14
Especially those located in major cities.
Ovum
November 2003
24
those with branch structures, through a single deal15. Our estimates suggest that
this set of customers represents between 10% to 15% of the total business
market. The ubiquitous incumbent can reach all of the sites of such customers.
However, its rivals cannot reach the small or remote sites without either
generating a loss or setting an uncompetitive price
• until very recently competitive local carriers and ISPs have reported substantial
problems in buying partial private circuits (or half circuits) at cost oriented prices
and at levels of granularity in terms of bandwidth which match their requirements.
Despite these problems, there is a general view among those interviewed for the
study that the (relatively) new obligations on incumbent operators to supply partial
private circuits, bitstream DSL services and unbundled local loops should address
these problems.
We recommend that the European Commission should:
• review the supply position for partial private circuits to see whether fixed
incumbents are now supplying partial private circuits at cost oriented prices in all
members states
• take no further action in relation to the corporate sector until such time as it is
clear that current measures are failing.
2.3 Competition in narrowband fixed services
Inspection of Figure 2.2 and analysis of the value chain indicate that:
• competition in the long-distance and international narrowband markets of the EU
is growing steadily towards satisfactory levels following liberalisation. In the UK
the incumbent now holds significantly less than half of this market. Other
countries which liberalised more recently are on track to reach a similar position
in time.
• competition in the local calls market is significantly weaker in virtually all member
states than competition in the long-distance markets and
• there is relatively little competition in the supply of narrowband access services –
especially in the mass markets of consumers and SMEs.
• except for the long distance calls markets, incumbents hold market shares of
between 80% and 100%
• new entrants in the long distance calls market typically pay 60% to 80% of their
turnover to incumbents for the use of their networks.
Based on our research we see little prospects of any increase in the level of
competition in the mass narrowband access market - although we see excellent
prospects for CATV operators supplying subscriber access to narrowband call
15
For example, by offering a VPN service or through an outsourcing deal.
Ovum
November 2003
25
services as part of triple play offerings. There are four main barriers to increases in
competition here:
• there are substantial economies of scale in the provision of access networks
which make it difficult for infrastructure based rivals to become cost competitive
• the old monopoly operator still has the advantages of incumbency. Entrants must
make substantial marketing and sales expenditure to attract customers.
Measures which require number portability have lowered the cost of attracting
customers but costs remain substantial
• the margins on narrowband access services are small or non-existent.
Traditionally monopoly operators subsidised line rentals out of call revenues.
Even after re-balancing margins on mass narrowband access remain unattractive
in most members states
• a high proportion of the cost of building an access network is sunk. Once
investments are made they cannot be re-used for other purposes. This raises
substantially the risks of market entry.
Wholesale line rental
This analysis has led several EU regulators to impose obligations on the fixed
incumbent to supply narrowband lines to rivals on a wholesale basis through a
wholesale line rental (WLR) product. This means that the incumbent is required to
rent the narrowband access lines, and associated features provided at the local
switch, to a rival at a wholesale price. This price may be cost oriented (as in the UK
and Ireland) or set on a retail minus basis (as in Denmark). The purchaser then
combines WLR with:
• a carrier selection service;
• cost based call origination and call termination products and
• use of its own facilities,
to provide a complete narrowband service to the customer.
There is considerable disagreement amongst those interviewed during the study in
relation to the merits of WLR. Proponents argue that:
• the barriers to competition in the supply of narrowband access are permanent
and insuperable;
• WLR gives all mass market customers a choice of total telephone service
packages from different operators rather than simply a choice of long distance
operators; and
• in the long term, WLR will enable NRAs to lift retail price control on incumbent
operators completely.
Opponents argue that:
• WLR significantly reduces incentives for investment in alternative infrastructure
especially by CATV operators who are considering investments in triple play
Ovum
November 2003
26
offerings. WLR is a national product and is not confined to those areas where
there are limited prospects for increased infrastructure-based competition;
• WLR is expensive to implement and requires micro regulation to specify the
functionality of the WLR products which the incumbent should offer and to ensure
adequate access to the OSS of the incumbent. BT, for example, claims to have
spent several tens of millions of euros to date in implementing WLR in the UK;
• WLR is inefficient in its use of network resources16. As Figure 2.3 shows, a local
call from a WLR customer can use up to five times as many switches as an onnet call or 2.5 times as many switches as a call between two interconnected
infrastructure operators;
• WLR is aimed at a market where prospects are poor and revenues shrinking; and
• WLR delivers little in the way of economic benefits. Call prices are already at
competitive levels and the scope for greater efficiency in retail activities is limited.
For example, we are told that the scope for reducing the costs of billing and debt
collection by utilities who do joint billing of gas, electricity and telephony services
are small.
Given the balance of arguments listed above, we recommend that Member States
investigate further whether WLR is effective in removing barriers to competition or
whether it acts to deter infrastructure competition.
16
NRAs could overcome this problem by requiring incumbents to offer service
providers end to end conveyance over their networks at cost oriented prices. But such
an obligation – which is similar to the UNE-P requirement in the USA – would
substantially undermine incentives for alternative infrastructure investment e.g. by
CATV operators wanting to offer triple play services.
Ovum
November 2003
27
Figure 2.3 The network inefficiencies of WLR - worst case scenario
Switches
used
On net local call
Incumbent’s
local
switch
A
A
Local call between interconnected infrastructure operators
Incumbent’s
local
switch
WLR call
A
Incumbent’s
local
switch
A
Rival’s
local
switch
A+B
B
Incumbent’s
trunk
switch
WLR
operator’s
switch
B
C
2A + 2B + C
The likely impact of a market decline
Another important barrier to competition in narrowband services is the prospects for
this market segment. The narrowband fixed services market is starting to contract,
after many years of strong growth:
• PSTN traffic in Denmark, Finland, Japan and Korea is already falling – typically
at 10% to 20% per annum; and
• we expect traffic in other Member States to follow this trend shortly.
There are a number of explanations for this new trend. The most likely factors
include:
• the move to DSL access. We believe that much of the recent strong growth in
PSTN traffic was generated by Internet dial-up access. This is now rapidly
migrating to DSL access;
• email is often substituted for fax and, to a lesser extent, for voice calls; and
• users are switching from fixed to mobile services for voice calls. This is a
relatively small effect so far in the European Union (when compared with the
USA) as we discuss further below
• users are beginning to switch voice traffic from circuit switched to VoIP based
services offered by new entrants. This has had a major impact in Japan where
traffic on the fixed circuit switched network fell 20% last year. VoIP services,
especially those based on PC to PC communication and presence functions,
could have a very substantial impact on voice-traffic in the future.
The poor prospects for narrowband services have a number of implications for
competition in this sector:
Ovum
November 2003
28
• falling traffic volumes mean that the incumbent’s fixed costs are recovered from a
shrinking number of minutes. This in turn means higher unit costs for incumbents
and so higher interconnect prices and lower margins for their rivals. However,
falling traffic volumes mean higher unit costs only to the extent to which the size
and capacity of the legacy network is maintained. We can expect that incumbents
will manage to migrate at least part of their stranded equipment to broadband and
mobile markets where traffic and capacity needs are growing;
• the incumbent will have spare capacity on its circuit switched network and will
probably price to use the spare capacity, while maintaining revenues17. This will
put additional competitive pressure on rivals; and
• we will not see new entry into this declining market.
Such consequences of reduced traffic volumes are not amenable to regulatory
intervention and should not be interpreted as an indication of market failure. It is
important for NRAs to recognise this and to consider carefully what intervention, if
any, is required, if and when hard pressed service providers seek regulatory relief as
a result of such changes.
Competition between fixed and mobile platforms
Given the poor prospects for infrastructure competition in the narrowband fixed
markets, and the problems associated with wholesale line rental, we might expect the
mobile networks to offer the greatest competition to the fixed incumbent for the supply
of basic voice telephony.
In this respect a number of points are worth considering:
• the evidence of competition between fixed and mobile platforms for basic voice
telephony in the EU is mixed. The proportion of mobile only households can be
as high as 30% in some member states as shown in Figure 2.4. In general a high
proportion of mobile only households reflects limitations on the availability of fixed
network services and mobile subscription fills this gap. Only in Finland do we see
really strong displacement of fixed line access by mobile subscription. At the
same time a recent study18 indicates that there is, as yet, relatively little
competition between fixed and mobile networks for voice traffic in the EU with
only about 1% of such traffic transferring from the fixed to the mobile network
each year
• some observers believe that we will see much greater competition between fixed
and mobile operators in the EU once the mobile operators start to roll out 3G
networks. These will provide approximately four times more capacity for voice
calls than existing 2G networks and significant improvements in call quality. Filling
them with voice traffic taken from the fixed operators is one way to start earning a
17
For example, by charging a relatively high cost per call and a low or zero cost per
minute.
18
Mobile and Internet substitution; Enders Analysis, 11/02.
Ovum
November 2003
29
return on the substantial investments required. It would appear that Hutchison
3G is adopting such a strategy in the EU countries where it is starting operations
• call termination charges for fixed operators are regulated at cost based prices
while, in most member states, mobile call termination charges are significantly
above cost. This generates a transfer of money from the fixed to the mobile
industry which should, at least in theory, strengthen the mobile industry’s ability to
compete with the fixed network operators
• there is much stronger competition between fixed and mobile operators in the
USA, where the volume of fixed network long distance minutes is falling at 6%
p.a., largely as a result of mobile substitution. This is partly as a result of very
different price packaging in the USA and partly as a result of lower premiums over
the prices charged by fixed operators for long distance calls. An obvious barrier
to such competition in the EU is the fact that the fixed incumbent owns one of the
main mobile operators in many Member States. It is clearly not in the interests of
such players to increase competition between fixed and mobile networks for voice
traffic. In contrast, in the USA, where the main competition is for long distance
minutes, five of the six largest mobile operators are independent of long distance
carriers or are owned by ILECs19.
We analyse this issue in more detail in Chapter 7, where we look at possible
divestiture by the incumbent, and in Chapter 8, where we look at measures to
increase cross-platform infrastructure competition.
19
The sixth, Sprint PCS, is quoted separately on the US financial markets as a
tracker stock. Sprint PCS reports separately from the rest of Sprint on its
performance to the investor community. This then leads to changes in the price of
the tracker stock. But the shareholder must trade Sprint shares as a whole and
cannot trade Sprint PCS tracker stocks on their own.
Ovum
November 2003
30
Figure 2.4 The proportion of mobile only households in the EU
35%
30%
% households mobile only
Finland
25%
20%
15%
10%
5%
0%
88%
90%
92%
94%
96%
98%
100%
102%
% households fixed and/or mobile
Source: EU Telecoms Services Indicators 2002
2.4 Mass broadband services – the state of
competition
Demand for mass broadband fixed network services20 is growing rapidly. By the end
of 2003 we expect the number of such services to pass 25 million in the EU and by
the end of 2007 to reach 90 million. In comparison there are just under 200 million
fixed access lines currently in use in the EU.
When considering barriers to competition in this market we need to distinguish
between the mass market of consumers and small businesses and the corporate
market of large businesses. We consider the mass market below, and the corporate
market in Section 2.2.
Many of the barriers to competition which exist in the mass market for narrowband
services do not exist in the corresponding broadband market. For example:
• the broadband market is a relatively new one and many of the advantages of
21
incumbency do not exist;
20
For this report we define as broadband any service offering more than 128kbit/s
downstream to the customer.
21
However, remaining advantages include the ownership of a ubiquitous network, a
well-known brand name, ownership of a large pre-existing customer base, detailed
knowledge of market behaviour and customer characteristics, and massive
Ovum
November 2003
31
• the European Commission introduced the local loop unbundling Regulation
2887/2000 at the end of 2000. This requires incumbents to offer shared or full
use of local loops at cost oriented prices and so helps transfer many of the
economy of scale advantages in the provision of access networks from the
incumbent to its rivals; and
• at the same time, the retail market price for broadband access is several times
higher than that for narrowband access.
So it is not surprising that the level of competition in the supply of broadband access
is higher than in the supply of narrowband access. But our research indicated that
there are still significant barriers to competition in the supply of mass broadband
services. As the analysis of this section and Sections 2.5 and 2.6 show:
• economies of scale make it difficult for alternative infrastructure operators to
compete with the fixed incumbent in the supply of broadband access. At the same
time the sunk cost nature of the investments required raises the risk of market
entry. Only CATV operators, who can exploit economies of scope between the
provision of TV and telecommunications services, are in a good position to
provide infrastructure based competition to the fixed incumbent in the supply of
mass broadband products
• incumbents have exploited their first mover advantage over those rivals who must
rely on them for input to their own services. In many cases they have delayed
local loop unbundling and in some cases offered wholesale DSL products after
their retail DSL product. But these advantages are no longer especially relevant.
Looking forward the main issues lie elsewhere
• the high upfront investment required of local loop unbundlers is the main barrier
to competition through local loop unbundling.
Figures 2.5 and 2.6 show the supplier position for broadband services at the end of
June 2003.
investment in access networks, ducts and buildings which the entrant must duplicate
to compete on equal terms.
Ovum
November 2003
32
Figure 2.5 The supply of broadband services in the EU at 31/3/03
Country
CATV retail using
CATV network
Austria
Belgium
Denmark
Finland
France
Germany
Greece
Ireland
Italy
Luxembourg
Netherlands
Portugal
Spain
Sweden
UK
330,000
400,000
157,304
63,000
312,707
45,000
0
4,000
0
1,071
842,000
263,000
404,473
156,400
1,098,000
Entrant retail
using own
facilities (excl
CATV)
0
0
13,425
1,000
0
0
1,640
3,000
192,200
250
0
0
0
138,285
8,000
Total
4,076,955
357,800
Entrant retail
Entrant retail
using ULLs using incumbents
wholesale DSL
Incumbent retail
using its own
DSL
Total
9,750
5,407
59,790
65,000
62,933
250,000
359
924
154,019
670
170,413
295
9,749
14,095
5,600
38,000
80,761
9,258
9,000
717,000
0
0
196
400,000
10
84
18,526
308,514
118,300
540,000
160,000
551,814
321,645
206,000
1,320,000
3,600,000
0
4,249
975,000
8,290
544,249
90,722
985,053
341,000
550,000
537,750
1,037,982
561,422
344,000
2,412,640
3,895,000
1,999
12,369
1,721,219
10,291
1,556,746
372,543
1,707,789
768,080
2,201,600
809,004
2,239,649
9,658,022
17,141,430
Source: ECTA Scorecard June 2003
We can see that:
• the level of competition in the supply of broadband services varies considerably
across the EU. In some countries, like France, Germany, Italy, Luxembourg and
Spain, the incumbent supplies over 80% of broadband services at the retail or
wholesale level. But in other countries, like Austria, Greece, Ireland, the
Netherlands, Portugal and the UK, the incumbent supplies less than 50% of such
broadband services.
• the main source of infrastructure competition to the fixed incumbent comes from
CATV operators. Across the EU as a whole, CATV operators supplied 26% of
broadband services at March 2003. In countries like Austria, the Netherlands,
22
Portugal and the UK, they supplied more than 50% of broadband services .
• only a very small proportion of broadband access pipes are supplied in other
ways by competitive access providers. Of the 5.9 million broadband connections
in October 2002 only 213,000 or 4% fell into this category. The bulk of these are
fibre connections supplied to businesses in city centres. Given the financial
difficulties which many of these operators face we do not expect this proportion to
grow substantially over the next few years.
• local loop unbundling is relatively little used but its use is growing rapidly. In
October 2002, the European Commission’s eighth implementation report
indicated that 2.5% of broadband services were supplied using ULLs. According
to ECTA this proportion had grown to 3.4% by March 2003 and to 4.7% by June
2003.
22
Despite passing less than 60% of households in three of these countries - Austria,
Portugal and the UK.
Ovum
November 2003
33
• incumbents sell a significant proportion of their broadband DSL services
wholesale. Across the EU as a whole, incumbents have sold one wholesale DSL
service for every five they have sold retail.
Comparison of the statistics produced by ECTA also suggests that the fixed
incumbents are strengthening their grip on the mass broadband market. In Q2 of
2003 alternative operators installed 8% more broadband services while incumbents
supplied 17% more23.
Figure 2.6 The level of infrastructure competition in the supply of broadband
100%
Incumbent retail
using its own DSL
90%
80%
Entrant retail using
incumbents
wholesale DSL
70%
60%
ULLs
50%
30%
own facilities (excl
CATV)
20%
CATV retail using
CATV network
40%
10%
UK
Lu
xe Ital
y
m
Ne bou
r
g
th
er
lan
ds
Po
rtu
ga
l
Sp
ain
Sw
ed
en
Au
str
Be ia
lgi
u
De m
nm
ar
k
Fin
lan
d
Fr
an
c
G
er e
m
an
G y
re
ec
e
Ire
lan
d
0%
Source: ECTA Scorecard June 2003
In analysing barriers to competition in the supply of mass market broadband access
we consider first the prospects for infrastructure based competition to the fixed
incumbent (Section 2.5) and then the measures which NRAs might take to promote
competition in retail broadband supply by imposing obligations on fixed incumbents to
offer wholesale products (Section 2.6).
23
Wholesale and retail
Ovum
November 2003
34
2.5 Mass broadband – infrastructure based competition
The EU versus major trading rivals
At the moment, the EU lags behind a number of its main trading rivals in the
development of broadband. As Figure 2.7 shows:
• the EU lags behind Japan, Korea and the USA in broadband penetration, but is
expected to close the gap over the next two years;
• price does not appear to be a factor which is suppressing demand in the EU;
• the proportion of broadband services supplied by alternative infrastructure
suppliers24 in the EU is significantly below that in Japan, Korea and the USA;
• the potential for the EU to increase broadband infrastructure competition to the
level enjoyed by these trading rivals is limited by the lower proportion of homes
passed by CATV networks when compared with the USA or Korea. Even so,
statistics from a recently published OECD report25 suggest that take up of
broadband cable modems per hundred homes passed is currently much lower in
the EU than in the USA, Korea or Japan; and
• the great success of unbundled local loops for broadband services in Japan is
based on pricing. NTT supplies unbundled local loops at $1-60 per month and
cheap dark fibre to rivals. These prices appear to be set based on historic costs,
rather than any economically efficient forward looking pricing standard. This leads
to prices which are significantly below replacement cost and which remove
incentives for further infrastructure investment. We would not recommend such
an approach in the EU.
Comparisons between countries suggests that there are limits on the extent to which
the EU can increase broadband infrastructure competition to match its trading rivals.
26
However, it is clearly important, given the merits of infrastructure competition , that it
should attempt to do so. We review the prospects below.
24
We include services provided by unbundled local loops in this category.
25
Broadband and Telephony Services over CATV networks, OECD, August 2003.
26
As discussed in Chapter 8.
Ovum
November 2003
35
Figure 2.7 Comparisons between the EU and major trading rivals
Measure
EU
Japan
Korea
USA
0.9%
4.0%
10.3%
1.2%
6.8%
10.9%
10.4%
25.5%
32.9%
3.3%
7.6%
13.4%
%of broadband services provided by:
alternative infrastructure service providers
ULL based infrastructure service providers
26%
3%
25%
40%
50%
~0%
57%
3%
Total %
28%
65%
50%
60%
$20 to$40
1.5%
$18
0.7%
$25
3.1%
$40 to $50
1.5%
47%
27%
70%
97%
Broadband penetration/00 population
1/01
1/03
1/05
Retail price of basic DSL service:
$ per month
% of GDP per head per year
Households passed by CATV networks as %
all TV households
Source: Ovum research; OECD
Competition from CATV networks
As Figure 2.5 shows, CATV operators currently offer the strongest competition to the
fixed incumbent in the supply of broadband services over alternative infrastructure.
This competition is possible largely because of the economies of scope which CATV
operators enjoy in the “triple play” provision of TV, telephony and fast Internet
access27. Moreover, the development of VoIP now offers the prospect for CATV
operators to add voice services to broadband Internet access and television services
at relatively low cost. Accordingly, we might expect CATV operators to provide strong
broadband competition to the incumbent through triple play offerings over the next
few years. As Figure 2.8 shows, CATV networks pass more than 50% of households
in ten of the fifteen Member States and, in many cases, the proportion of households
28
which subscribe to CATV network services continues to grow .
In addition e.biscom has shown that, in the absence of CATV operators, viable
infrastructure competition based is possible using other conventional technologies.
e.biscom for example uses a mix of optical fibre and unbundled local loops rather
than coaxial cable connections to offer triple play services in the major cities of Italy.
27
We define fast Internet access as access at downstream speeds in excess of 128
kbit/s.
28
In contrast to PSTN access lines where demand is stagnant.
Ovum
November 2003
36
Figure 2.8 CATV network reach in the EU Member States
Member state
Austria
Belgium
Denmark
Finland
France
Germany
Greece
Ireland
Italy
Netherlands
Portugal
Spain
Sweden
UK
% homes passed
by CATV network
53%
100%
71%
59%
35%
83%
na
78%
1%
94%
60%
29%
65%
50%
% growth in Subscribers
last 4 years
2001 (m)
17%
3%
8%
14%
58%
21%
na
42%
na
7%
190%
1200%
9%
80%
1.2
3.8
1.1
1.0
3.4
21.8
na
0.6
0.1
6.2
1.1
0.5
2.1
3.6
Source: Ovum estimates based on OECD statistics
The prospects for future growth of the CATV operators are mixed. In some countries,
like Spain and Portugal, we have seen significant growth in the number of CATV
subscribers in the last four years. But this is from a relatively small base. In other
countries, like Belgium and the Netherlands, CATV networks are long established and
market penetration is high. So the prospects for growth are limited. In other
countries, like the UK, financial problems limit the prospects for growth. The financial
performance of the two main CATV operators in the UK - NTL and Telewest – are
open for inspection. It is clear from such inspection that:
• once the CATV network is built, it is possible to generate good profits from
supplying bundles of broadband and narrowband services to the mass market.
NTL and Telewest both operate at EBITDA margins in excess of 20% and project
margins comparable with those generated by incumbents within two to three
years; but
• the substantial capital investment required to build the networks has so far
generated negative returns and is unlikely ever to generate a market rate of
return.
We conclude from this analysis that:
• we will see growth in some member states in both the coverage footprints and
subscriber take up for CATV services.
• even in countries where financial difficulties limit expansion of the coverage
footprint, CATV operators will exert significant competitive pressure on the fixed
incumbent.
There is a further barrier to future infrastructure based competition from cable
networks. In several countries the major part of the CATV infrastructure remains in
the ownership of the fixed incumbent. We consider this issue in detail in Chapter 7 of
our report.
Ovum
November 2003
37
Competition from new technology networks
Are there other possible sources of access network infrastructure competition?
Alternative technologies - such as fixed wireless access and use of electricity power
lines– are possible options which, if successful, could create strong infrastructure
based competition to the fixed incumbent. Our assessment of the prospects of these
two technologies are very different:
• powerline communications (PLC), in which telecommunications signals are
carried over the electricity distribution network from the local transformer to the
customer premises, has been a promising but little used technology for 10 years.
For many years limited reliability and interference problems have prevented
successful mass deployment of PLC. Now the recent liberalisation of the
electricity sector, advances in chip technology, and better control over radio
interference have combined to generate strong renewed interest in PLC. With
175 million power lines into customer premises29, the potential impact of PLC on
infrastructure-based competition is very substantial.
• fixed wireless access, like PLC, has a lot of theoretical attractions and may be
deployed by fixed incumbents as the lowest cost way of serving rural customers30
with broadband access. But whether fixed wireline access will provide competitive
local carriers with the infrastructure they require is highly debatable. A recent
Ovum study suggests that the unit costs of fixed wireline access mean that it is
simply not cost competitive with ADSL access. Figure 2.9 illustrates.
For the moment there is considerable debate about whether the second generation
PLC equipment which is now available can safely be deployed or whether it will
interfere to an unacceptable degree with existing radio communication services in the
1.6 to 30 MHz bands. The European Commission is leading a consultation process to
try to resolve these difficulties and might issue a Recommendation on PLC
deployment in future. For the moment however the position is highly uncertain. There
are (at least) three possible outcomes:
• PLC becomes a major new technology for providing broadband access and offers
a powerful new source of infrastructure based competition to the fixed incumbent
and the CATV operators.
• PLC interference problems continue to restrict deployment.
• PLC interference problems are solved but PLC still does not offer viable business
models to the power companies and their partners for mass broadband
deployment. In many ways the power companies using PLC face similar upfront
costs in providing backhaul from electricity substations to a core network as local
loop unbundlers face when backhauling from a main distribution frame. As we
discuss in Section 2.6 such upfront costs have played a major part in suppressing
demand for unbundled local loops.
29
Over 200 million in the enlarged EU.
30
Where local loops are too long to allow use of ADSL.
Ovum
November 2003
38
We conclude that:
• given PLC’s potential as a source of infrastructure based competition it is
important to resolve the interference problems which PLC equipment currently
faces.
• uncertainties over the future impact of PLC on infrastructure based competition
make it important to consider regulatory measures to strengthen competition in
the supply of broadband services in addition to any impact which PLC will have.
Figure 2.9 The unit costs of mass broadband technologies
Technology
Total cost
Cost per line
Notes
DSLAM: $940,000
$205
Approximate costs based on a
9,000-line DSLAM
$275
200-line DSLAM
$2,100
200-line SDH box
$510
Approximate costs based on
1,000 customers
$1,010
1,000 customers
$2,700
1,000 customers
Fixed line
Large ADSL DSLAM
CPE: $100
Small ADSL DSLAM
DSLAM: $35,000
CPE: $100
SDH over optical fibre
SDH multiplexor:
$20,000
CPE: $2,000
Wireless
Licence-free FWA
Base station: $10,000
CPE: $500
3.5GHz FWA
CPE: $1,000
26GHz FWA
CPE: $2,000
Source: Ovum
Competition through alternative ownership models
There is another possible model of supply which involves the use of conventional
technologies but novel forms of ownership. Under such models:
• a building developer or the community owns the access network which serves a
business park, large apartment block or new housing estate; and
• this owner allows various service providers to connect to its access network at a
point of interconnection and serve the tenants on a competitive basis.
Such ownership models exist in a number of EU Member States. We can see no
need for regulatory intervention to either promote or constrain such models, providing
Ovum
November 2003
39
that the tenant has the right to choose his or her service provider (so as not to distort
competition). We do not know whether tenants have such rights in all Member States.
We recommend that Member States should review the rights of tenants to choose
their service provider where a landlord owns the local access network to which wide
area network services are connected.
Infrastructure competition - the impact of state funding
The European Commission recently approved the use of structural and government
funds to support broadband rollout in less favoured areas. Such funding is intended to
minimise the prospects of a digital divide emerging between the different parts of the
EU. In approving such funding, the European Commission set out conditions under
which such funds can be applied31. For example:
• funds should support programmes which are part of coherent development plans;
• funds should be targeted at areas that would “be neglected under free market
conditions”;
• funding should only be granted where there is a competitive environment and
where the incumbent’s tariffs have been fully re-balanced; and
• funding should be technology neutral.
Despite these guidelines, virtually all respondents to our study remain concerned that
state funds will lead, and already are leading, to distortions of competition. They claim
that the EU guidelines on use of state funding are being ignored. All support the idea
of using government funds to minimise the digital divide between urban and rural
communities, as long as:
• such funds are used only in areas where market led investment will not take
place. Funding in other areas reduces incentives for alternative infrastructure
investment. Operators claim that local authorities are already breaching such
32
constraints in places like Dublin, Norwich (UK), Amsterdam and Stockholm;
• such funds are used in ways which do not distort competition. Alternative access
network operators are concerned that government authorities will issue requests
for proposal which, by their size and nature, rule out such operators from making
a competitive bid, effectively handing the contract to the incumbent; and
31
Guidelines on Information Society and Telecommunications Infrastructure, March
2003.
32
Where fibre access is planned using State funds
Ovum
November 2003
40
• government authorities, in selecting successful bidders, take account of dynamic
as well as static economic efficiencies33 Current EC guidelines explicitly prohibit34
such considerations.
We recommend that the European Commission investigate these complaints further
to see if they have substance and, if so, to develop more detailed guidance for
Member States to follow on funding of broadband rollout and other ECNS platforms.
2.6 Mass broadband – service based competition
Introduction
In carrying out their market analysis NRAs will need to consider:
• whether infrastructure based competition35 provides sufficient competitive
pressure on fixed incumbents so as to maximise the benefits of competitive
supply of mass broadband access services; and
• if not, what wholesale broadband products a fixed incumbent should supply in
order to maximise sustainable and effective competition.
It is clear from Figures 2.5 and 2.6 that there is considerable variation in the
competitive supply of broadband services at the retail level and that fixed incumbents
will need to supply wholesale products under ex-ante regulation in some, but not
necessarily all, member states. So in this section we consider how best to resolve
these two issues.
Market definitions and findings of SMP
The answer to the first question is closely linked to the market reviews in which the
NRAs are currently involved and in particular the review of the wholesale unbundled
access market (Market 11) and wholesale broadband access market (Market 12). It is
inappropriate for us to offer any guidance on this review process. But based on our
discussions with interviewees in the course of our study we note the following points.
First there is considerable debate on the precise definition of sub-markets, especially
within Market 12. In particular there is debate on:
• the lower limit on the downstream speed of a service before it is excluded from
broadband services. Some authorities propose a lower limit of 128kbits/s; others
33
i.e., they should consider the position five years from now. If they choose the
incumbent now, will they effectively be opting for a monopoly in perpetuity with all the
loss of dynamic efficiency that would entail?
34
e.g., Section 5 on Tendering Processes
35
Supplemented by the activities of local loop unbundlers which Regulation
2887/2000 enables
Ovum
November 2003
41
256kbit/s. In addition, some operators argue that broadband and narrowband
services are, in certain circumstances, in the same market.
• whether symmetric and asymmetric broadband services are in the same or
different markets. A number of NRAs have ruled that these two service types are
in separate markets.
• whether symmetric broadband services are in the same market as the wholesale
terminating segment of leased lines (Market 13 identified in the
Recommendation).
• whether broadband access36 and broadband conveyance are in the same or
different markets.
• whether wholesale services for business and residential customers are in the
same or in different markets.
• whether there is a single national market for the supply of wholesale products or
whether the geographic areas in which CATV operators offer service are in a
different market from those areas where they are absent.
Secondly the answer to the question of whether the fixed incumbent is considered to
have SMP in these markets, once defined, could vary considerably across the EU.
• in some Member States, it is reasonably clear that there is a strong case for an
SMP finding in relation to Market 12.
• in others37, where the fixed incumbent holds less than 50% of the retail
broadband market, the case requires careful consideration. Any SMP
assessment is complicated by the fact that, while CATV operators in these
countries compete vigorously at the retail level, they rarely offer wholesale
broadband products. There are a number of reasons put forward by the CATV
operators for this lack of wholesale supply:
-
it is technically more difficult for CATV operators, who run a shared access
network, to supply wholesale products than it is for incumbents who operate
local loops dedicated to specific customers;
-
CATV operators believe that they can generate more revenue by selling
broadband Internet access on a retail basis alone. Selling wholesale might
lead to more customers, but it would weaken their brand and reduce
opportunities for cross-selling TV and voice telephony services; and
36
Defined as carriage from the end user to the local exchange or RCU.
37
e.g., Austria, the Netherlands, Portugal and the UK.
Ovum
November 2003
-
42
CATV operators typically only cover a part of a Member State, while potential
ISP customers normally offer services nationally38. This lack of national
coverage means that a CATV operator would need to sell at prices below the
incumbent’s wholesale offering to generate demand.
• some argue that the lack of wholesale products from CATV operators means that
the incumbent has SMP, even when CATV operators supply the major part of the
retail broadband market. Others argue that such a market is contestable and that
CATV operators would begin to supply wholesale services if the fixed incumbent
raised its price significantly. If contestability is established, the case for ex ante
regulation is weaker.
• these issues are inherently tied with an assessment of the appropriate manner in
which to address self-supply and captive supply in the context of the new
regulatory framework.
What products should the fixed incumbent offer?
If the fixed incumbent is found to have SMP in the wholesale broadband market(s),
what services should it be required to offer? Currently incumbents offer three broad
types of wholesale product39, as illustrated in Figure 2.109, namely:
• IP level wholesale DSL service in which the purchaser buys the incumbent’s
retail DSL product at a wholesale price set by the market or by the NRA on a
retail minus basis. Normally, the pricing structure and the functionality of this
product reflect that of the retail product. The bitstream might be delivered to the
ISP via a managed or unmanaged IP network and may or may not be bundled
with global Internet connectivity.
• bitstream DSL service. The incumbent still provides a DSL service and
backhaul40 to the purchaser’s point of presence. However, the purchaser is able
to specify the balance between upstream and downstream speed, the contention
ratio for the backhaul, and the location of the point of interconnect within the
incumbent’s ATM network. Such a product enables the purchaser to differentiate
its offering from that of the incumbent and to offer different retail price structures.
• unbundled local loops. The purchaser rents the local loop (or uses the high
frequency spectrum within the loop) from the incumbent and provides DSL
service and backhaul itself. Such a product is currently offered by all operators
with SMP in the national interconnect market, under EU Regulation 2000/2887.
38
Although in some cases the length of the local loops may make DSL supply
impossible.
39
There are many variants on the three categories described above.
40
We show ATM backhaul in Figure 2.3. But IP/MPLS and Gigabit Ethernet
backhaul is also possible.
Ovum
November 2003
43
Which of these product types should an SMP operator be mandated to supply on an
ex ante basis?
Mandating IP level service
NRAs are currently deciding whether to require operators with SMP in relevant
markets to supply IP level services. In so doing, they need to consider whether there
are grounds, under the new framework, to mandate supply on an ex-ante basis or
whether competition law offers adequate remedies to deal with any incumbent who
refuses to supply or supplies at a price which creates a margin squeeze on an
independent ISP which wants to purchase such a service. Such consideration is
required by the guidelines set out in the European Commission’s Recommendation of
February 2003. These suggest that NRAs should only introduce ex ante measures
where existing competition law remedies are inadequate to deal with a market failure.
If NRAs do rely on competition law remedies then it is important that any fines
imposed for breach of competition law are proportionate to the gains that the SMP
operator has made as a result of the breach. Unless this is the case, competition law
is unlikely to be a sufficient constraint on the operator’s behaviour.
Figure 2.10 The main wholesale broadband services
DSL access
PC
Backhaul
MDF
DSL
Local
loop
Splitter
ATM
DSLAM
Managed or
unmanaged IP
network
ISP
ULL
Bitstream access 1
Bitstream access 2
IP level wholesale
Mandating supply of unbundled local loops
Incumbents with SMP in a relevant market similar to Market 11 are currently required
to supply unbundled local loops. NRAs will need to decide whether to maintain,
amend or withdraw this mandate as part of the market review process. However,
given the low current level of use of unbundled local loops we need to consider
whether more can be done to promote take-up without distorting competition.
Our research indicates that the main reason for the poor take-up of unbundled local loops
is the high up-front investment required. An unbundler will typically pay €70,000 to
Ovum
November 2003
44
€100,000 in collocation and backhaul costs before it can start to provide service to any of
the customers attached to the main distribution frame in connection with which these
costs are incurred. In other words it appears that the natural monopoly in the provision of
local loops extends back into the incumbents network – at least from the MDF to the local
exchange41. Given that:
•
each MDF serves 3,000 local loops on average in the EU42;
•
each DSL service might generate €500 per annum from a consumer or €3000 per
annum from a small business; and
•
each DSL customer might have a three year life before changing supplier,
it is vital that the unbundler is able to win to a substantial share of customers if it is to
recoup its initial €70,000 to €100,000 expenditure. Such investment is clearly risky and,
in today’s post-dotcom climate, few investors are willing to take such risks especially for
mass market roll out of broadband services. It is possible that take-up of unbundled local
loops will increase if and when investor confidence in the telecommunications sector
returns. However, many of those interviewed in the course of our study believe that this
problem is permanent.
A few respondents, particularly the local loop unbundlers, also argue that unbundled local
loop take-up is low because incumbent operators have delayed implementation and
raised unbundlers’costs unnecessarily (e.g. by insisting on secure collocation cages
rather than allowing co-mingling of equipment). However, there is a general view that
such problems can be, and are being, overcome as the NRAs and the operators work
their way through the operational problems which entrants face when renting local loops,
so as to develop workable reference offers from the incumbents.
Despite these problems, the number of unbundled local loops is growing rapidly and our
research indicates that there is significant demand to rent local loops from firms wishing
to supply end users with substitutes for 2 Mbit/s leased lines. At the same time there are
measures which NRAs could take to reduce the upfront cost for such firms. In particular
they could:
• mandate the option for the local loop unbundler to commingle its equipment with
that of the incumbent rather than use secure collocation cages. According to
ECTA such commingling, which can significantly reduce collocation costs, is so
far mandated in only seven43 of the fifteen member states;
• require the incumbent to rent cost oriented backhaul at an appropriate level of
disaggregation which matches both the unbundler’s requirements and the
41
But it is very difficult for an NRA to establish the boundary of such monopoly ex
ante. It will depend on a wide variety of factors and will probably change over time
42
There is considerable variation in the size of main distribution frames. Many
terminate less than 1,000 loops. Others terminate many tens of thousands of loops.
Not all of these loops are short enough to be used for broadband service
43
Belgium, France, Ireland, the Netherlands, Portugal, Spain and the UK
Ovum
November 2003
45
incumbent’s ability to supply in a cost effective quantity. Given the incumbent’s
existing transmission capacity between the MDFs and the core network it is highly
likely that the incumbent can supply such backhaul at a significantly lower cost
than the unbundler.
This analysis prompts us to make two recommendations:
• that incumbents with SMP in Market 11 be required to offer co-mingling of
equipment to local loop unbundlers.
• that incumbents with SMP in Market 11 be required to offer cost-oriented
backhaul at an appropriate level of disaggregation to local loop unbundlers.
Mandating bitstream services
The possibility of mandating the provision of bitstream products44 has been a subject
of considerable debate. Some operators argue that:
• mandating local loop unbundling45 is as far as regulation should go to promote
competition in broadband services;
• mandating bitstream access will allow purchasers to win substantial broadband
revenues whilst the incumbent takes all the investment risk. This, in turn, will
reduce the incumbent’s investment incentives, slowing broadband rollout; and
• mandatory bitstream access reduces the investment incentives for operators,
such as local loop unbundlers and CATV operators, who provide greater
infrastructure-based competition than those who buy bitstream access.
We take a different view. We believe that:
• the incumbent enjoys substantial economies of scale in providing DSL access.
For example, it is able to purchase DSLAMs in bulk at prices which are 50% or
more below the prices paid by its smaller rivals. Mandatory bitstream allows the
purchaser to enjoy these economies of scale also;
• bitstream access reduces significantly the upfront investment which the purchaser
must make, when compared with local loop unbundling, thereby removing one of
the major barriers encountered in relation to local loop unbundling;
• bitstream access is significantly better at promoting competition in broadband
markets than IP level wholesale services. Bitstream access enables
differentiation in terms of quality of service and price when compared to IP level
wholesale service; and
• the case for bitstream-type services is likely to strengthen in the future. As
incumbent operators introduce VDSL access and push the electronics required to
provide service closer to the customer, the business case for local loop
44
45
By incumbent operators with SMP in relevant markets.
With associated supporting measures.
Ovum
November 2003
46
unbundling becomes weaker46 and the incumbent’s dominance in the access
network strengthens.
Of course, the arguments about investment incentives are important ones. They
disappear, however, if the bitstream product is priced so as to reward the incumbent
fully for the risk it is taking. We discuss this point further in the next sub-section.
We therefore consider that it would be a reasonable and proportionate NRA response
if operators with SMP in the relevant markets were required to offer bitstream DSL
access products for which there is reasonable demand and in areas where the
incumbent offers retail DSL access.
If these suggestions are accepted, they will, in Member States where the incumbent
has SMP in the supply of wholesale broadband services, offer a migration path from
service-based competition to infrastructure-based competition in which the service
provider:
• buys IP level wholesale service as Step 1;
• buys bitstream service as Step 2;
• migrates to unbundled local loops as Step 3; and
• finally migrates to building its own access network facilities as Step 4.
As a service provider moves from Step 1 to Step 4 along the path:
•
it moves from service-based to infrastructure-based competition;
•
it takes more of the investment risk itself;
•
it pays less to the incumbent for its access components; and
•
it is able to offer products with greater functional and price differentiation.
Supporters of this hierarchy of services argue that, by using this migration path, a service
provider can reduce its investment risks substantially. It can build its customer base using
Step 1 or Step 2 products and then migrate to Step 3 or Step 4 products once it has
established a critical mass of customers in an area. For this approach to work, certain
conditions must be met. For example, it is important that:
all of Steps 1 to 3 are available to service providers4748. In many Member
States, this is not the case;
a)
46
As the number of points of presence required for a given level of coverage
increases by an order of magnitude and the up-front cost to the local loop unbundler
of reaching these points of presence increases by a similar amount
47
And they are available on a long term basis.
48
Service providers should be able to obtain IP level wholesale service using
competition law remedies
Ovum
November 2003
47
b)
charges by the incumbent for migrating the service provider’s customers from
one step to another are set at the efficiently incurred cost level. Little
attention has been paid to this aspect to date;
c)
incumbents are able to implement the migration between different products
without leaving the end-user without service. Several of the respondents to
our study have identified this as a major problem; and
d)
the relative prices of the incumbent’s offerings for Steps 1 to 3 are set
correctly. Pricing a lower step product too low deters migration towards
infrastructure competition. Pricing it too high restricts competition
unnecessarily.
To date, against the background of a poor investment climate, there is no clear
picture as to how the migration path will work in practice. However, we believe that it
is important that such a policy is initiated in order to promote competition over
infrastructure in the broadband area.
We therefore recommend that NRAs ensure that conditions (a) to (c) set out above
are met by the incumbent’s offerings. We discuss pricing further below.
The pricing of mandatory wholesale products
There is detailed work now going on at the EU and Member State level to consider
the most appropriate pricing of wholesale broadband products. So again we do not
offer any prescriptive guidance on this subject. But this is clearly an area of great
importance to NRAs and operators of all kinds. In the course of our interviews, we
spent considerable time discussing this issue. Through these discussions we
identified six constraints on the pricing of mandatory wholesale products which we set
out below. It is important to note that these six factors constrain pricing; they do not
determine prices.
In setting prices for mandated wholesale products, we believe that NRAs should be
guided by the following six pricing constraints.
Constraint 1: the need to maintain pricing relativities. We can think of the three
wholesale product types and the corresponding end user DSL service – the four steps
of the previous section - as a price and cost stack as shown in Figure 2.11. The NRA
could, at least in theory, set mandatory wholesale prices for all three products on
either a cost plus or retail minus basis. If it uses the cost plus approach, it will also
need to decide how to allocate the common costs across the cost stack. If it uses the
retail minus approach it will automatically allocate the common costs to the unbundled
local loop component of the stack.
Ovum
November 2003
48
Figure 2.11 The price and cost stack for broadband products
Retail
price
Retail costs
IP level
wholesale
price
IP backhaul costs
?
?
Bitstream
price
Common
costs
ATM backhaul costs
?
DSLAM and collocation costs
ULL
Price
?
Local loop costs
Constraint 2: the existence of competition law remedies. Service providers can
use competition law to lower the incumbent’s wholesale prices. For example,
aggrieved service providers might claim that an incumbent is operating a margin
squeeze in its pricing of IP level wholesale, and possible bitstream, services. Such
action or, if the fines are big enough, the threat of such action, could be enough to put
a ceiling on the prices of these products.49 This is an argument for NRAs not to
mandate pricing for IP level wholesale products.
Constraint 3: unbundled local loop prices are currently set in all Member States
at cost oriented levels. The economically efficient pricing standard for unbundled
local loops is relatively uncontentious. In contrast with bitstream services, local loops
are technologically legacy products for which demand is stable, in which little new
investment is required, and which are not subject to rapid technology change. As
such, an NRA can establish a price which provides the incumbent with full recovery of
50
its LRICs knowing that it will not affect the incumbent’s future investment behaviour
whilst, at the same time, given a reasonable ‘build or buy’signal to entrants wishing to
offer access services to customers.
Constraint 4: the need to preserve incentives for the incumbent and its rivals to
invest in the rollout of broadband infrastructure. This is not a particularly
important constraint on bitstream pricing if a retail minus approach is used. The
incumbent can then increase the return it earns on its bitstream product investment
simply by raising its retail prices and rivals can follow suit. However, the constraint is
of vital importance if the NRA decides to adopt a cost oriented approach to bitstream
pricing. Here products have two components:
49
Equal to the incumbent’s retail price less its avoided retail costs.
50
More accurately its long run incremental costs plus a mark-up for common costs.
Ovum
November 2003
49
• the local loop for which pricing is relatively straightforward - as set out in
Constraint 3.
• the new technology component of the product, such as DSLAMs, splitters and
ATM backhaul, where another approach is required.
A cost oriented price which gives an incumbent (and its infrastructure-based rivals)
the right incentives to invest in products based on new technology and new plant will
need to take account of:
• investments in new products which fail. A significant proportion of new ventures
by telecommunication operators fail. The costs of these investments must be
recovered from the products which succeed. The NRA should not consider only
the rate of return on the successful products. Figure 2.12 illustrates this point for
the pharmaceutical industry, where R&D risks are especially high. However, the
investment risks of incumbents may be lower than those of comparable new
infrastructure operators where incumbents can rely on existing ubiquitous
networks, a established brand and user base and where they may succeed in
leveraging market power from adjacent markets.
• the asymmetric risk which incumbents and new operators face when they invest
in new technology products in competitive markets where the sunk investment is
a high proportion of the total cost. In such markets additional demand leads to
increased prices and capacity in the industry then increases to limit price rises. A
decrease in demand however, does not lead to less capacity. Accordingly, prices
fall rapidly. This asymmetric pricing behaviour as demand fluctuates means that
firms are wary of investing unless they receive a higher price than that which
simply recovers cost of supply at expected demand levels. If the incumbent’s
broadband wholesale products are price capped at this latter price, the incentives
to invest in broadband rollout are limited. At the same time, the incumbent’s rivals
are keen to buy inputs from the incumbent rather than to invest in their own
infrastructure.
• the rate of improvement in the price/performance of new technology components.
For example, we have already seen the development of a second generation of
DSLAMs with substantially improved price/performance. For an incumbent to
invest in a first generation asset, it must be confident that it can recover the
purchase price. In a competitive market, the prices it can charge to do this fall
over time, as the price/performance of the technology improves. As a result, the
incumbent needs to charge prices when the first generation technology is
introduced that compensate for the lower prices that it can charge as the asset
reaches the end of its life.
Ovum
November 2003
50
Figure 2.12 New product profitability in the pharmaceuticals industry
After tax present value in $m (1990 prices)
1200
1000
800
600
400
Average R&D cost
200
0
1
2
3
4
5
6
7
8
9
10
New products by decile
Source: H Grabowski “Price and Profit control, New Competitive Dynamics and the Economics of Innovation in the
Pharmaceutical Industry”, in A Towse (editor) “ Industrial policy and the Pharmaceutical Industry", 1995
In a competitive market these effects can lead to a doubling or trebling of a simple
LRIC price51 to reach a price level at which the rational incumbent will invest. Setting
such a price is clearly challenging. The challenge is even greater if, as seems likely,
the market is only partially competitive. For example the incumbent may not be price
constrained by the existence of better price/performance technology in these
circumstances. So the impact of the last of the three factors listed above is
weakened and the premium required over the simple LRIC price to create investment
incentives is reduced.
In the USA the FCC recognised this point in its recent report and order on the
unbundling obligations of the ILECs:
“The effect of unbundling on investment incentives is particularly critical in the
area of broadband deployment, since incumbent LECs are unlikely to make the
enormous investment required if their competitors can share the benefits of the
facilities without participating in the risk inherent in such large scale capital
52
investment.”
One pragmatic approach to allow for these effects is to shorten asset lives and/or
increase the return on capital employed to a level above the incumbent’s weighted
average cost of capital. However there is, as far as we know, no accepted rule of
51
Calculated in a risk free scenario where economic and technical depreciation are
identical.
52
FCC 03-06 released on 21 August 2003.
Ovum
November 2003
51
thumb for determining the precise percentage changes to these parameters which
should be made.
Constraint 5: The need to ensure maximum pricing freedom at the retail level.
The broadband market is a new one and a key factor in maximising demand is the
ability of operators to price at the retail level with the greatest possible freedom.
Subject to fact-specific intervention on the basis of competition rules,53 operators
need:
• to be able to charge different prices at different times. For example, introductory
offers may stimulate demand which in turn stimulates production of broadband
content. This then increases the value of broadband to customers and allows
operators to raise prices without losing them; and
• to charge different prices to different customer groups - offering different price
packages to high and low volume users, to those who want high bandwidth and to
those who are attracted principally by the “always on” characteristic of DSL
services.
Retail minus pricing constrains such price differentiation more than a cost oriented
approach to setting wholesale product prices does.
Constraint 6: the need to prevent excessive end user prices in the long term. If
prices for wholesale broadband products are set on a retail minus basis and the
incumbent has SMP in retail broadband markets, there are no substantial constraints
on the incumbent setting monopoly prices in the long term. In such circumstances,
the NRA must consider introducing retail price controls. However, if wholesale prices
are set on a cost plus basis and there is a sufficient level of competition or a credible
threat of market entry, there are strong constraints on the retail prices which the
incumbent can charge for broadband services and intrusive retail prices controls are
not required.
Dynamic pricing for wholesale broadband services
A number of NRAs - notably the CTRC in Canada and OPTA in the Netherlands have set dynamic prices for unbundled local loops. Under such schemes, the initial
54
price of unbundled local loops is set at a level significantly below cost and gradually
increased, typically over a five year period, until it reaches a cost-based level.
Such a scheme represents a straightforward form of entry assistance which promotes
the quasi-infrastructure investments which local loop unbundlers must make.
Moreover, it does so in a manner which makes it clear that prices will rise. As a result,
there should be no significant danger that such an offer will encourage inefficient
53
for example SMP operators will need to ensure that they do not create a margin
squeeze for their rivals. They also need to avoid foreclosure through the strategic
targeting of discriminatory pricing practices.
54
In the case of the Netherlands, the starting price in 1999 was the historic cost to
KPN and the final price the current cost of providing local loops.
Ovum
November 2003
52
entry. In countries where CATV competition is weak, NRAs might justify this measure
on the grounds that it promotes quasi-infrastructure-based competition and the
dynamic benefits which such competition generates (as described in Section 8.3).
Does such a measure work? There is some evidence to suggest it might. Figure
2.13 indicates that there is a reasonable correlation between the price currently
charged for ULLs and the current take up of ULLs55 . The Netherlands, excluded
from the trend line, is an outlier where the current ULL prices are high by EU
standards, but so too is ULL take-up. One possible explanation for this outlying
position is that use of dynamic pricing has stimulated ULL take up. However, we
advise caution in interpreting this situation. The data do not correspond to the same
point in time; Japanese take-up is probably influenced strongly by the supply of cheap
dark fibre backhaul by NTT; there are also many other factors which might influence
take up in the EU. At the same time, OPTA’s measures were introduced when the
market was in its infancy and may not be appropriate in today’s market conditions.
However, the evidence available from the Dutch situation does suggest that the idea
is worth exploring further.
Figure 2.13 LLU prices v take-up
45.00%
40.00%
Jap
% of DSL using ULLs
35.00%
30.00%
y = -0.03x + 0.397; R squared = 0.82
25.00%
20.00%
Fin
NL (excluded from
the regression line)
.
Dk
15.00%
10.00%
It
5.00%
D
A
S F B
0.00%
0.0
2.0
4.0
6.0
8.0
10.0
E
12.0
P UK
14.0
16.0
Monthly rental for ULL ($)
55
We have excluded Ireland from this plot because the market there is still at a very
early stage of development.
Ovum
November 2003
53
2.7 Public WLANs
Public WLAN services are currently an area of considerable interest in supplier
activity. They also provide functionality which powerful fixed and/or mobile operators
might wish to integrate with mobile services. So we consider them in this chapter.
Demand for private WLANs is strong. Organisations of all type, and even consumers,
benefit from use of private WLANs to avoid the cost and inconvenience of cabling.
But demand for public WLANs (often referred to as Wi-Fi) is still far from certain.
The appeal of public WLAN services is clear. In theory end users can establish high
bandwidth connections to IP networks at a wide range of “hotspots” such as cafes,
railways stations, hotels and airports. They can then use their laptop PCs or PDAs to
access the Internet, emails and corporate network services.
But in practice the market is highly fragmented. Different hotspots are operated by
different service providers; no single service provider offers widespread coverage;
and roaming between the public WLANs is limited and very variable in its
functionality.
Right now countries like the USA, Japan and Korea lead in the rollout of public
WLAN. EU member states, with the notable exception of Sweden, are lagging behind.
Ovum forecasts, as shown in Figure 2.14, that revenues from public WLAN services
in the EU will grow to around €0.5 billion per year by 2008. This compares with total
current annual telecommunications revenues in the EU of around €300 billion.
Figure 2.14 Ovum forecasts of global public WLAN service revenues
The attitude of mobile operators towards public WLANs is mixed. Many of them see
such services as a threat to their future 3G services and revenues and some have
Ovum
November 2003
54
lobbied the authorities to place restrictions on public WLAN services; others have
launched public WLAN services themselves. Some observers also suggest that, in
member states where the fixed incumbent owns the major mobile operator:
• the supply of public WLANs will be less competitive because two of the obvious
suppliers of such services are in common ownership; and
• the incentives for discriminatory deals between the fixed incumbent and its mobile
subsidiary in the development of services which integrate public WLAN services
with mobile services are substantial.
Given the likely scale of public WLAN services over the next five years and the
current embryonic state of the market we believe that it would be premature for
regulatory authorities to intervene in this market.
We are concerned about the scope for discriminatory deals between the fixed
incumbent and its mobile subsidiary. We discuss this point further in Chapter 7.
2.8 Infrastructure vs service based competition
Many of those interviewed in the course of our study raised the issue of what kind of
competition is possible in each of the main segments of the ECNS market. In
combination they made five main points:
• infrastructure based competition is inherently superior to service based
competition where it is viable. In particular infrastructure based competition gives
incentives for the fixed incumbent to innovate in network technologies, product
functionality and pricing packages which do not exist when the incumbent faces
only service based competition;
• the experiences of the last five years suggest that there are significant limitations
on where infrastructure based competition is viable, especially in the delivery of
fixed network services;
• in markets where infrastructure based competition is not viable it makes sense to
introduce measures designed to promote service based competition;
• such measures can provide a stepping stone to infrastructure based competition
in that they enable a service provider to build a customer base and revenues from
services constructed by renting network components from the incumbent before
migrating customers to its own facilities. Such a process lowers the entry risks for
service providers; and
• such measures can also undermine incentives for infrastructure based
competition and, if priced wrongly, even remove incentives for investment by both
the incumbent and its rivals in new infrastructure and technology upgrades.
These points are highly relevant in the fixed network services markets but they apply
more widely to ECNS markets as a whole. Accordingly, we consider them in detail in
a separate chapter – Chapter 8.
Ovum
November 2003
55
2.9 The move to next generation networks
There is now strong consensus that the next decade will see a gradual move from
circuit switched to next generation IP networks across the developed world. This
transition should lead to networks which can, at lower unit costs, deliver a wide range
of multimedia, broadband data, and voice services.
This migration has begun. EU incumbent operators, such as BT and Telecom Italia,
have now largely replaced their trunk circuit switched network with next generation
switches using ATM and/or IP technologies. The pace of the migration, the nature of
the applications which will run on next generation networks, and the way in which
operators will charge for these services, at either the retail or wholesale level, is far
from certain. However, it is reasonable to expect that, within a few years, the markets
will develop to a point where a substantial proportion of traffic is generated by sites
connected directly to a next generation network.
It is also hard to predict what level of competition will develop in the supply of these
next generation networks. In the period prior to the dot.com crash, large numbers of
potential entrants developed business plans to launch next generation networks. We
might reasonably expect, as investor confidence returns and incumbents offer a
range of broadband access products at the wholesale level, that we will see
significant competition in the supply of next generation networks. The principles set
out in Figure 2.15 below are based on such a competitive scenario.
These developments prompt us to ask two important questions:
•
what are the regulatory issues which need to be resolved now in relation to Next
Generation Networks and which issues can be left to future studies when the form
of Next Generation Networks developments will be clearer?
•
do NRAs need to set out guidelines now, along the lines of Figure 2.15, relating
to the regulation of next generation networks, so that operators can invest with
confidence?
Ovum
November 2003
56
Figure 2.15 Possible principles for regulating next generation network
services
1. NRAs should forebear from regulation of innovative new services until there is clear
evidence that such forbearance is harmful to the development of competition
2. NRAs should not prohibit any operator from exercising first mover advantage on
innovative next generation network services
3. NRAs should allow all operators, including incumbents, considerable freedom in setting
the retail prices for next generation network services – both in terms of bundling and low
initial prices.
4. Incumbent operators should not be required to offer unbundled services from their next
generation networks unless and until there is substantial demand for such products.
5. When incumbent operators are required to offer next generation network services at the
wholesale level at regulated prices, NRAs should take full account of demand uncertainties
and technology risks in setting the price of supply.
In general, those interviewed as part of our study believe that:
•
the issues raised in our study on Next Generation Networks are very important
and will require study at some point; but
•
that it is too early in the development of Next Generation Networks for us to try to
address these issues now.
Based on these findings, we recommend that the European Commission should
monitor closely the regulatory issues which arise as a result of use of Next
Generation Networks.
The following issues are of special importance:
• what is the likely pace of implementation of next generation networks within the
core networks and access networks of incumbent operators in the EU? Such
migration is now starting to gather pace for both fixed and mobile operators. It is
important that regulatory authorities in the EU understand the speed of this
migration.
• to what extent will the incumbent’s use of IP networks for Internet access raise
the cost of its rivals in a manner which reduces competition without compensating
cost efficiency and/or public welfare gains? We highlight this issue in Section 5.4
of our report.
• to what extent will the migration to next generation networks create new control
points? As the recent Devoteam report56 points out, the architecture of next
generation networks means that there are new control points such as the ability to
56
Regulatory Implication of the Introduction of Next Generation Networks and other
New Developments in Electronic Communications, European Commission, May
2003.
Ovum
November 2003
57
bundle service offerings, access to APIs and interoperability of soft switches
through which a dominant operator can exercise market power. It would clearly
be damaging to regulate on an ex ante basis to prevent abuse of this power while
the technology is in an embryonic stage. However, it is important to identify the
control points and monitor their use by powerful players.
• is there a need for NRAs to impose new interoperability conditions on next
generation network operators beyond the basic set which applied to traditional
circuit switched networks? This issue relates back to the control point issue set
out above.
• where, if at all, in the next generation hierarchy should interconnection be made
mandatory? The use of softswitches to control calls and deliver applications will
generate a new set of issues for NRAs to tackle. In particular it is important to
consider how an operator with SMP would supply each of the wholesale services
listed in the European Commission’s Recommendation57 using a next generation
network and where interfaces with other networks are required.
• for which next generation services is ex ante regulation required in the short
term? The logic behind European Commission’s Recommendation suggests that
any voice call termination service offered by next generation networks may be
regulated on an ex ante basis.
• what form should such ex-ante regulation take and which of Articles 9 to 13 of the
Access and Interconnect Directive are appropriate? In particular is cost
orientation required?
• how should any regulated cost oriented prices, such as call termination charges,
be calculated for a next generation network? There are a number of challenges
here:
-
the cost structure of next generation networks is uncertain. But we do know it
will be characterised by a much higher proportion of fixed costs than for a
circuit switched network and that the variable costs which are attributable to
voice are likely to be small;
-
the cost of carrying a voice call through the network will depend upon the
quality of service which is provided. In a circuit switched network there is one
basic level of quality of service. But in a next generation network there is a
range of quality of service levels which are possible, each with different cost
implications; and
-
the link between the duration of the call and the costs generated is less clear
cut in a next generation network than in a traditional circuit switched network.
• should the cost of voice call termination for a fixed incumbent be established as a
blended average of the unit cost of circuit switched call termination and the unit
cost of next generation network call termination? The principle of technology
neutrality suggests that market definitions are based on services and not
underlying network technologies. This in turn suggests that a fixed incumbent
57
Commission Recommendation 2003/311/EC.
Ovum
November 2003
58
which terminates (say) 50% of its call on a next generation network and the other
50% on its legacy circuit switched network, should have its call termination
service regulated at a price which reflects the unit costs averaged across the
networks. This averaging would help to deal with a possible rise in unit cost on
the circuit switched network as call volumes on this network fall.
• how do NRAs apply the principle of non-discrimination to a next generation
network? Given the various quality of services which next generation networks
offer there is, prima face, a case for interconnect agreements with and between
next generation networks which specify quality of service levels.
Ovum
November 2003
3
59
Mobile network services
3.1 Introduction
This chapter examines barriers to competition in the supply of mobile network
services:
• in Sections 3.2 to 3.5 we start by looking at the current structure of the industry
and the market trends which the industry faces. In particular, we consider the
growing importance of data services, the impact this has on supply chains and
the potential barriers to competition which then arise;
• in Sections 3.6 to 3.19 we then go on to examine the various barriers to
competition which we identified through our analysis.
3.2 Current industry structure and trends
The industry structure in mobile markets in the EU varies between voice and data
services, as shown in Figure 3.1. In most Member States:
• the industry is based around three to five competing network operators; and
• the mobile operators have spent the last few years of the 1990s integrating
forward in the supply chain, buying up independent service providers and
independent mobile phone retailers so that now very few are left and the industry
structure is concentrated.
Figure 3.1 Industry structure in EU mobile markets
Content
Provision
Application
developers
Content
providers
Content
Aggregation
Network
Operations
Service
Provision
Retail
Mobile network operators
Portals
Independents
MVNOs
voice
data
Source: Ovum 2003
Some of the operators have also bought operators outside their home country in an
effort to reap economies of scale. Vodafone has carried this off most successfully – it
now has holdings in an operator in most of the EU Member State markets. The other
main players who have attempted this horizontal integration are T-Mobile, FT/Orange,
O2, and KPN. See Figure 3.2.
Ovum
November 2003
60
Figure 3.2 International positions of operators in Europe
Operator Group
Main countries of current operations in Europe
Vodafone
Belgium, France. Germany, Greece, Ireland, Italy, Netherlands, Portugal, Spain,
Sweden, UK
Orange
Austria, Belgium, Denmark, France, Italy, Netherlands, Portugal, Switzerland,
UK
T-Mobile
Austria, Germany, Netherlands, UK
O2
Germany, Ireland, Netherlands, UK
KPN
Belgium, Germany, Netherlands,
TeliaSonera
Denmark, Finland, Norway, Sweden
Over the last decade there have been repeated attempts in different countries to set
up mobile virtual network operators (MVNOs)58 who buy wholesale airtime and retail it
to their subscribers. Examples include Netsystem and Sense Communications in
Norway, Talkline in Germany and Carphone Warehouse in the UK with its Value
Telecom brand. These have largely failed because of high charges for wholesale
airtime and difficulties in dealing with operators which led the small MVNOs to run out
of money before agreement was reached. Today, the only truly successful MVNO in
Europe is Virgin Mobile, which is, in any event, run as a joint venture between Virgin
and T-Mobile in the UK.59 It has over 2 million subscribers in the UK market.
However, even here there are problems, in that the two owners are suing each other,
each trying to force the other to sell its share of the venture.
As the mobile industry started to develop data services in the late 1990s it expanded
to include application developers, content providers and content aggregators. At first,
the mobile operators thought that they could take on many of these functions
themselves and started integrating backwards up the value chain. However, over the
last two years, they have retreated from this position. Most run their own portals but
prefer to deal with a community of application developers and content providers. They
also provide access to third party portals through their WAP services.
Over the last three years, there have been a number of new 3G licences issued. As a
result, many EU Member States will see one completely new operator enter the
market during 2003 or 2004. Although this will reduce concentration in the industry
slightly, it will not fundamentally change industry structure (not least because many
commentators expect a number of the smaller 2G operators to exit the market or form
alliances of some nature with entering 3G operators in the medium term).
58
We define MVNOs to include enhanced service providers such as Virgin Mobile.
59
We note that we understand that Virgin’s recent attempts to conclude other joint
ventures in other countries appear to have stalled.
Ovum
November 2003
61
3.3 The value chain for data services
The value chain for the delivery of data services is longer and more complicated than
the value chain with which the mobile industry is familiar. Figure 3.1 illustrates. It is
also different from that for fixed network broadband services. Mobile operators are
trying to differentiate their data services from simple high-speed Internet access by
also providing a wide range of content services.
There are some major differences between the value chains for serving the consumer
market and the business market in relation to data services.
The consumer value chain
At present, most of the effort is going into developing and launching consumer
broadband services. The basic consumer value chain is:
• a wide array of content providers. These range from large media companies
such as Sky, through film companies to the highly fragmented ring-tone providers
and – in fact – any web content;
• a large number of application developers, who sell their applications to the
network operators either outright, or on a hosted basis. These range from
developers of mobile games through to those providing email and personal
information management. The mobile operators are also active in this area for
applications built around messaging, such as instant messaging, picture
messaging and chat services;
• content aggregators, including:
-
the mobile operators themselves who are mostly running their own mobile
portals (e.g. Vodafone’s Live) as walled gardens;
-
some existing portals which are offering mobile versions as part of their “triple
play” strategy, e.g., MSN, Virgin and AOL;
-
third party aggregators who are selling some content on a hosted basis
mostly to smaller mobile operators;
• the mobile network operators;
• service providers. The mobile network operators mostly take on this role
themselves, although there is increasing interest in the possibility of data MVNOs
from companies involved in web content (e.g., Uboot.com); and
• the retailers.
In addition to this basic value chain, additional players are involved in new areas of
activity, notably:
• m-commerce – here there are additional players such as payment processors
and banking networks; and
• wireless marketing – here there are also additional players including wholesale
SMS providers and mobile advertising networks such as 24/7.
Ovum
November 2003
62
The business value chain
For business data services the value chains are simpler because most of the content
is located within the business (email, intranet, business applications) or on the
Internet. However, there is a lot of development from the major enterprise software
vendors (e.g., Microsoft, Oracle, SAP) to make corporate applications work properly
over mobile networks.
There are also developments in the area of service provision. Many of the large IT
service companies – like IBM and EDS - now provide mobile services as part of larger
deals that they have with their clients for supplying, integrating and supporting
business applications.
3.4 Major players and their market shares
Voice services
The biggest market in the mobile industry is that for retail voice services, accounting
for some 85% of industry revenues across the EU.
Supply in this market is concentrated in the leading one or two operators, who
account for 70 to 90% market share (by subscriber numbers) in most countries. This
is often the incumbents’mobile subsidiary and the local Vodafone company. The
remaining operators typically hold 10 to 20% share each. Figure 3.3 illustrates.
When looking across countries we can see that the geographic expansion of the
major operators means that the industry is also becoming concentrated at a
European-level, with Vodafone accounting for around 30% of voice traffic and other
groups active across parts of Europe (excluding Telefonica and TIM here) accounting
for 43% of such traffic. See Figure 3.4.
Ovum
November 2003
63
Figure 3.3 Share of voice traffic of operators in EU Member States
Operator share December '02
Austria
Belgium
Denmark
Finland
France
Germany
Greece
Ireland
Italy
Netherlands
Norway
Portugal
Spain
Sweden
Switzerland
UK
1
45%
54%
46%
63%
49%
43%
41%
57%
47%
42%
69%
48%
56%
47%
63%
27%
2
30%
30%
28%
29%
36%
37%
29%
40%
34%
27%
31%
29%
25%
38%
20%
25%
3
20%
16%
14%
8%
15%
12%
29%
3%
19%
12%
23%
19%
16%
17%
25%
4
5%
5
13%
8%
1%
0%
11%
24%
9%
Top 2 operators
75%
84%
74%
92%
85%
80%
70%
97%
81%
68%
100%
77%
80%
84%
83%
51%
Source Ovum 2003
Figure 3.4 European share of voice traffic of major mobile operator groups
Group
Vodafone
Orange
T-Mobile
TIM
O2
Telefonica
KPN
Subscribers
Dec 02 (m)
88
50
40
28
19
19
13
Total Europe
Share
29%
17%
14%
9%
6%
6%
4%
300
Source Ovum 2003
Narrowband data
Retail narrowband data60 currently accounts for some 15% of mobile industry
revenues in Europe. This is dominated by SMS, which constitutes approximately 90
to 95% of narrowband data revenues. The remaining revenues come from business
use of GPRS for email and Internet access. Picture messaging is expected to
expand data revenues, but the service is too new for the effects to be quantifiable.
60
Services offering data speeds of less that 64 kbit/s. Includes GPRS and SMS.
Ovum
November 2003
64
Shares of narrowband data traffic broadly reflect those of voice, with only minor
differences where an operator prefers to concentrate on particular segments (e.g.,
Europolitan in Sweden focuses more on the business segment).
Broadband data
The market for broadband mobile data services61 in Europe will start during 2003,
with tens of launches of third generation UMTS networks based on W-CDMA
expected before the end of the year. However, at present it is in a fledgling state and
its share of the industry’s total revenue is negligible.
Equipment and software supply
Generally, the equipment and software vendors supplying the players who deliver
services are regarded as being in a competitive market. However, Nokia’s position in
the European mobile markets is noteworthy. Its market share of mobile devices in
Europe is between 60% and 70%. It has been active in building the industry structure
for data services, setting up a large community of developers (Forum Nokia) and a
content aggregation service (Club Nokia). This is seen as a threatening move by
some operators, who worry that their data service business could be undermined by
it. Nokia insists that its only interest is to seed the market.
3.5 Main market trends
Overall
The number of mobile subscribers has started to saturate at between 70% and 90%
of the population in most Member States. The big drive for operators now is to
increase ARPU, through data services. However, many forecasters are projecting
CAGR for revenues per user at only a few % per year. This changes the market
conditions substantially, increasing pressure for lower unit costs.
Voice
Ovum expects voice ARPU to remain flat over the next five years. Declining call
termination revenues will be roughly offset by increased retail minutes and revenues
(and even an increase in retail prices).
There are 30% more mobile connections than PSTN lines in Europe, but mobile
originated voice still represents a small share of total voice traffic. Traffic seems to be
stabilising at about 15% of the total – at least in the UK. See Figure 2.2.
Some mobile operators are experiencing congestion in densely populated areas.
Spectrum trading between GSM operators could ease this. Many operators want to
61
Services offering data speeds of 64 kbit/s or more. Includes W-CDMA and EDGE.
Ovum
November 2003
65
use W-CDMA for voice capacity relief but cannot plan for this as the handset vendors
are not planning to launch cheap voice-centric W-CDMA phones in the next two
years.
Mobile operators are interested in generating revenue growth by attracting voice
traffic from the fixed network. On most GSM networks the business case for this is
uncertain because of the extra capital expenditure needed. Many operators will wait
until their W-CDMA network is launched before exploring this opportunity.
There is concern that the launch of W-CDMA networks will increase voice capacity
too much, leading to a collapse in voice traffic prices.
Narrowband data
Narrowband data has seen huge growth over the last few years with the rise of SMS.
Figure 3.4A illustrates. It shows how the growth of SMS has outstripped voice traffic
in the UK over the last four years with the ratio of SMS messages to voice minutes
increasing from 1 to 10 to 1 to 3.
Figure 3.4A The relative growth of SMS and voice traffic in the UK
60000
35%
30%
50000
25%
40000
20%
30000
15%
20000
10%
10000
SMS meesages (m)
Mobile voice mins (m)
SMS to voice mins (%)
5%
0
0%
1999/00
2000/01
2001/02
2002/03
Source: Oftel market statistics
The number of services is now broadening and such services are the catalyst for
significant change in the industry:
• SMS dominates narrowband data revenues and continues to show rapid growth.
It is developing into a sophisticated market with a variety of billing arrangements
(premium and reverse billing), the use of special short codes for special SMS
services and a supply chain that includes wholesale SMS providers.
Ovum
November 2003
66
• MMS was launched in Europe during 2002. The market is embryonic, but is
showing promising signs. We expect a sophisticated market to emerge with time.
• WAP services had a very bad start in 2001 but we understand that they are
quietly picking up in popularity and usage, as operators get better at providing
portals and users become more familiar with the services.
• GPRS access to corporate data is currently a small market, serving people with
laptops and PDAs, but it is growing gradually, as GPRS networks are stabilising,
coverage is improving and roaming is becoming more reliable.
• Most operators are developing, usually in collaboration with specialist content
providers, new narrowband services, including Instant messaging, chat, locationbased enhancements to WAP services, games downloads and on-line games.
Narrowband data operators are starting to take greater control of the appearance and
packaging of their services, by specifying their own phone software. Vodafone Live
and Orange SPV are the best examples, but we expect other operators to follow.
Broadband data
Broadband data services on mobile networks will become a feature of the European
landscape during 2003, with the launch of 3G W-CDMA networks. However,
following the collapse of the tech sector “bubble”, the vision of broadband multi-media
services carried on W-CDMA has lost credibility.
In this environment, operators have sought to delay their 3G launch. Some have
even surrendered their licences. Efforts are being made to improve the business
case, including network sharing arrangements, to minimise capex, and testing EDGE
(the GSM upgrade) for complementary coverage in some areas.
There is a shortage of reliable and independent information on the progress of 3G WCDMA, but we understand that it is limited. Operators are continuing to acquire sites
as fast as possible. However, Nokia and Ericsson (who account for around 70% of
the W-CDMA infrastructure market) have shipped only 10,000 to 15,000 base stations
each world-wide. We understand that most operators are buying Asian equipment,
since it is already developed and that European suppliers are missing a major
opportunity here.
Most operators will launch W-CDMA with a fairly small coverage area, and then build
out the network over a period first to meet their regulated coverage requirement and
to meet coverage and capacity demand. All operators will rely on dual-mode working
where phones roam onto a GSM network when outside W-CDMA coverage.
There is growing speculation that GSM was the last technology that will be rolled out
on a national scale. Operators will depend on EDGE for broad coverage for high
bandwidth services.
While W-CDMA is in its difficult launch phase there is plenty of noise from substitute
technologies also offering broadband wireless data services, including public WLAN
and so-called “4G”technologies, such as those offered by Flarion and Arraycom.
Ovum
November 2003
67
Although there is no standardisation yet, the level of interest in 4G is rising because it
appears to promise much higher bandwidths at lower levels of capex than 3G. A
number of European players are considering running trials during 2003/4. To do this,
they may need to “borrow” some of their 3G spectrum. There are already trials
running, notably in Australia, Korea and the US.
3.6 The impact of consolidation
There is general, but not universal, support for the thesis that there will be significant
consolidation in the EU mobile industry over the next five years. This thesis is driven
by a view of the economies of scale which are inherent in the industry. These are of
two types:
• economies of scale within a country are important. Small operators have
substantially higher costs per subscriber than large ones and find it hard to
generate profits. Figure 3.5 illustrates. This makes them take-over targets for
larger operators in the same country.
• economies of scale between countries are more debatable and more difficult to
realise. Mobile operators with a number of country-specific subsidiaries (e.g.,
Vodafone, Orange, T-Mobile) may have a significant cost advantage over other
operators with only a single operating unit (e.g., Bouygues, Wind). New
applications such as multimedia messaging are developed once and then
localised. Accordingly, the development costs are spread over four or five
operating units. This may make single operator unit businesses vulnerable to
take-over.
Greenfield 3G licensees are particularly vulnerable, when compared with existing 2G
licensees holding a 3G licence. A 2G operator migrating to 3G has a number of
advantages over a pure 3G licensee:
• there are substantial economies of scope between 2G and 3G networks;
• customer acquisition costs are lower. The 2G operator only has to migrate
customers from 2G to 3G. The 3G licensee has to acquire them from scratch;
and
• the 2G operator can roll out network faster and more cheaply by using existing
2G resources (sites, distribution channels and skilled staff).
Accordingly, we may see 3G licensees combining with small 2G operators and/or
surrendering/ selling their licence.
Within this changing industry structure there is a general view that:
• the level of infrastructure competition in the supply of mobile network services will
remain satisfactory in most EU Member States – even after consolidation – with
three or more viable mobile operators competing;
• competition law, including forward-looking undertakings imposed in the context of
merger reviews, provides a satisfactory mechanism for dealing with any
competition issues which may arise from consolidation in these Member States;
Ovum
November 2003
68
• there are a few small Member States, such as Ireland, where consolidation may
well lead to the creation of a duopoly. In such countries, special measures may
be required to preserve an adequate level of retail competition.
Based on this analysis, we conclude that:
• there is no need for the European Commission to institute ex ante EU-wide
measures to deal with competition problems arising from consolidation of the
mobile industry; and
• it is important that NRAs do not intervene to preserve non-viable mobile operators
who come to them seeking regulatory relief.
Figure 3.5 Economy of scale effects – EU mobile operators
60%
50%
40%
EBITDA %
30%
20%
10%
0%
0%
10%
20%
30%
40%
50%
60%
-10%
-20%
Market share
Source Ovum 2003
3.7 International roaming
The continuing high level of retail prices of international roaming service remains a
feature of the mobile markets in the EU. We have seen some moves to cut prices by
multinational mobile operators such as Vodafone, T-Mobile and Orange. But still
prices remain high. Ovum’s own calculation suggests that charges are typically four to
six times the network costs of carrying the call whilst, in comparison, retail mark-ups
on domestic mobile calls are typically in the 50% to 80% range. INTUG also carried
out studies to compare the prices charged for carrying a call from Country A to
Country B for a domestic subscriber and for a visitor using a roaming service. The
differences in price were very substantial.
Ovum
November 2003
69
DG Comp of the European Commission is currently investigating whether the prices
charged for international roaming services are excessive and whether their high level
reflects anti competitive conduct by the operators. The investigators are expected to
publish their findings shortly.
Only one of the respondents to our study - INTUG – raised this issue. But this does
not mean that the issue is unimportant. Given our sample:
• we would not expect the various operators and service providers interviewed to
raise the matter.
• NRAs often see the problem as a trans-national issue which is already under
investigation by the European Commission. So they see no need to comment.
3.8 Spectrum trading
There is considerable interest among mobile operators in the idea of spectrum trading
and strong theoretical arguments in favour of introducing it. However, those
responding to our study hold a wide variety of views in relation to the best approach
to spectrum trading. No clear pattern emerges.
We understand that DG InfoSoc has already commissioned a study to look at this
matter in detail. We, therefore, make no recommendations on this issue. However we
point to the following issues raised by our consultation:
• prevention of spectrum hoarding;
• dealing with windfall gains;
• the appropriate restrictions on cross use of spectrum;
• the criteria for defining eligible traders;
• dealing with the potential skewing effects on competition; and
• the impact of spectrum trading on barriers to entry, exit and expansion by mobile
operators.
3.9 Use of SIMLOCK
In our study we raised the issue of whether mobile operators lock-in customers
through use of long contract periods and SIMLOCK and so create barriers to
competition.
It is clear from the responses we received that this is not a current problem and that
no regulatory intervention is required.
Given the substantial handset subsidies which exist in the EU there is a good case for
permitting mobile operators to SIMLOCK terminals in relation to which such subsidies
exist – both for commercial reasons and to protect end-users against fraud and theft.
In almost all cases, the period of SIMLOCK is less then the minimum contract period.
In effect, the SIMLOCK process is simply a way of enforcing the minimum contractual
term. At the same time, contract periods – typically 12 months for packages involving
Ovum
November 2003
70
subsidised handsets - are generally considered reasonable. The danger is that the
SIMLOCK may be extended, to operate as a barrier to churn.
3.10 MVNO hosting
Stakeholders were asked whether there was a case for regulatory intervention to
enable MVNOs to negotiate satisfactory supply terms from their host mobile operator.
The Hong Kong regulator, OFTA, has imposed such requirements.
Based on the responses that we received, our analysis is as follows:
• mandatory hosting would require mobile operators to offer wholesale service at
regulated prices. This reduces the incentive for infrastructure investment and the
likely level of future infrastructure competition between mobile operators;
• it is important to preserve the current levels of infrastructure competition in the
mobile industry as much as possible – particularly given the prospects for
consolidation over the next five years;
• with the introduction of 3G networks, there are strong market incentives for
mobile operators to reach satisfactory commercial arrangements with MVNOs.
MVNOs who offer new services, content, applications and marketing approaches
to specific market segments are, at least in theory, an excellent way for mobile
operators to diversify the range of services offered on their 3G networks, fill them
with traffic more quickly, and earn a return on their investments;
• even if these market incentives fail to produce satisfactory agreements with
MVNOs, it is too early in the development of 3G services for regulators to
intervene; and
• it is difficult to develop a common regulatory approach to MVNOs, given their
different treatment across the Member States in 3G licence conditions. Some
countries, like Ireland, have explicit conditions dealing with MVNOs; others, like
Italy, prohibit MVNOs. Any attempt to change these conditions is likely to face a
legal challenge.
Based on this analysis, we conclude that the case for mandating mobile operators to
host MVNOs is weak. We recommend that the European Commission does nothing
in the immediate future to require mobile operators to host MVNOs, subject to the
extent to which future consolidation suggests that the level of infrastructure-based
competition may become insufficient.
3.11 Access for value added service providers
Should NRAs require mobile operators to provide access to Value Added Service
(“VAS”) providers? Our analysis, taking into account the views of respondents,
follows
• many of the arguments which apply to MVNOs apply with equal force to VAS
providers seeking access to the mobile operators’facilities
Ovum
November 2003
71
• there is a requirement in most Member States for mobile operators to supply
location information to VAS providers62;
• mobile operators argue that they have strong commercial incentives to open up
their networks to VAS providers as a way of increasing their revenue streams. At
the moment they are competing with each other to attract them onto their
networks; and
• mobile operators have another incentive to work with VAS providers. They are
under competitive pressure from mobile terminal suppliers (like Nokia) who have
formed direct links with VAS providers (as discussed in Section 7.3). This
increases the commercial incentives for mobile operators to enter into
agreements with VAS providers.
Based on this analysis, we conclude that there is no requirement for regulatory action
to mandate agreements between mobile operators and VAS providers.
3.12 Supply conditions for greenfield 3G operators
Some respondents to our study raised concerns about the terms of roaming
agreements with 2G operators. In addition, they expressed concern in relation to
obtaining access to the existing sites of 2G operators. The problem, in both contexts,
appears to relate to the terms of access, rather than an outright refusal to provide
access. For example, 2G operators are willing to site-share, but only at a price that is
significantly in excess of the estimates of cost made by those entities seeking access.
There are also major problems for all 3G licences in obtaining the required number of
base station sites to roll-out their 3G networks. Environmental and health issues
make new site acquisition difficult for all. For example, some local authorities have
made it clear that they will only consider joint applications for site-build. The coordinated roll-out which is needed to make such joint applications is difficult to
achieve, especially for operators who are attempting to gain first mover advantage
through early roll-out.
There are significant problems for site acquisition and sharing, especially for
greenfield 3G operators. However, the problems are complex and vary by Member
State. It is difficult to envisage an appropriate general regulatory remedy.
3.13 Restrictions on network sharing
Mobile operators responding to our study hold mixed views on the extent to which
they want to use network sharing. They agree that:
62
It is important to note that data protection legislation requires prior notice and
consent before location data can be processed and transmitted to a 3rd party.
Ovum
November 2003
72
• in the long term, they all want to provide their own infrastructure. This gives them
much greater control over the nature of the services they can offer; and
• network sharing can significantly reduce investment costs during 3G roll-out.
With these general findings in mind we can see little merit in preserving restrictions on
network sharing given that:
• sharing is mandatory in some circumstances anyway - to minimise environmental
impact;
• lifting restrictions on network sharing does not force mobile operators to share;
• mobile operators have strong incentives to build their own infrastructure in the
long term; and
• lifting restrictions reduces costs during 3G roll out and may also reduce the
probability of industry consolidation (as discussed in Section 3.6).
We recommend that the European Commission should examine whether it is practical
for the restrictions on network sharing which exist in many Member States to be
removed.
3.14 Roll-out conditions for 3G networks
In our study, we asked whether the roll-out conditions imposed on 3G licensees
constitute a barrier to competition. The argument behind the question was as follows:
• 3G licences were drafted, and operators bid for licences, at a time when demand
forecasts for 3G were far higher than they are now;
• these conditions may now constitute a barrier to competition. Licensees must
either meet the conditions by risking substantial investments or effectively
surrender their licences;
• conditions should be relaxed so that operators can be free to invest in response
to market demand; and
• this is the only way to maximise investment in 3G infrastructure and competition
in 3G services.
Our subsequent investigations suggest that this argument has little substance
because:
• in virtually all Member States, except Sweden, mobile operators tell us that the
roll-out conditions are not especially onerous. For example, in France, mobile
operators are only required to cover 70% of the population within 10 years.
• there is only a problem in Sweden, where the auction for 3G licences was
evaluated largely on the speed with which bidders proposed to roll out their 3G
networks. However, even in relation to Sweden, the severity of the obligations
can be reduced. The use of W-CDMA technology means that cell sizes
“breathe”. A fully loaded cell has a radius of less than 1km, but a lightly loaded
cell has a radius of up to 20km. Accordingly, there is scope to interpret the
Ovum
November 2003
73
coverage conditions flexibly and align the roll-out conditions which are specified in
the licence to match the current projections of market demand.
Assuming that such a flexible approach is possible, we conclude that there is no case
for EU-wide action to modify existing 3G roll-out conditions.
3.15 The use of walled gardens for data services
Many operators offer data services by providing access to a walled garden portal. In
many cases, users are able to enter a URL to escape and use other parts of the
Internet. However, in some cases (e.g., Orange’s all-you-can-eat WAP access
package) users are barred from general surfing unless they pay additional charges.
Where an operator has specified the mobile phone software it is easier for it to set up
these restrictions. Clearly, such practices can shut out third party service providers.
Accordingly, we investigated whether regulation is needed in response to these
restrictions.
On the basis of the comments that we received on during our study, we have made
the following analysis.
• it is clear that many mobile operators, especially the smaller operators, are
following an open approach to data and Internet services, rather than a walled
garden approach. For example, Radiolinja, in Finland, and Bouygues, in France,
have adapted an open approach (while their main rivals have implemented a
walled garden approach);
• the walled garden approach has value in helping to build user confidence in new
data services and in offering consumer protection. Mobile operators can make
sure that walled gardens services are easy to use, offer predictable prices and
minimise the risk of fraud;
• given the level of mobile retail service competition, there are strong incentives for
mobile operators to move beyond the walled garden approach, if that is what endusers want; and
• regulatory intervention on walled garden issues now could harm the development
of embryonic data services.
Based on this analysis, we recommend that NRAs should forbear from regulatory
intervention on this issue.
3.16 Dominance in the mobile terminal markets
Nokia has an important market position in the supply of mobile handsets in the EU
(with a 60% to 70% market share) and a major supplier of mobile network equipment.
It is, therefore, theoretically possible for Nokia to leverage its position in the handset
market by ensuring that its network equipment works better with its own handsets
than it does with those of other suppliers.
Ovum
November 2003
74
Our enquiries indicate that this concern is without substance. There is no evidence
from those who responded to our study that Nokia is in any way attempting such
leveraging. There is also a strong view that Nokia’s dominance of the EU terminal
and network equipment markets is now under threat from Korean and Japanese
suppliers.
3.17 Cost based call termination prices
Several of those responding to our study raised, unprompted, the issue of cost based
call termination prices for mobile operators. They argue that:
• mobile call termination rates in the EU are significantly above cost;
• there are no adequate market mechanisms to force these rates to cost-oriented
levels;
• mobile operators discriminate in the way that they charge for call termination, in
that they apply one rate to themselves and another much higher rate to rivals;
and
• these market failures constitute major barriers to effective competition – both
within the mobile industry and in competition between fixed and mobile operators.
They point to the following competition problems:
• fixed operators are unable to compete with mobile operators on equal terms for
voice traffic. Mobile operators can subsidise their retail prices for outbound calls
from the supernormal profits they generate on call termination. Fixed operators
cannot;
• fixed operators cannot compete with mobile operators to deliver voice traffic from
a corporate network to mobile terminals. A mobile operator typically offers the
corporate customer one rate for such termination, whilst charging its fixed rivals a
significantly higher rate; and
• large mobile operators offer on-net retail prices which are below the call
termination charge that they offer other operators. As such, they argue that the
large mobile operator discriminates in favour of its own retail business and
against its rivals in the supply of a product in relation to which, according to
European Commission analysis, it is dominant. In addition, low on-net prices are
one of the ways in which mobile operators differentiate themselves, and low onnet prices give large mobile operators, with higher subscriber populations, an
competitive advantage over their smaller rivals. Applying the non-discrimination
requirements to narrow the gap between off-net and on-net retail call prices would
reduce this competitive advantage.
Recent regulatory analyses by the European Commission and the UK’s Competition
Commission support these arguments. In addition, the European Commission is
currently conducting an investigation into the pricing practices of KPN in relation to its
corporate VPN (fixed and mobile) services. Finally, we note that Oftel has just
initiated a case under Chapter II of the Competition Act to investigate the second of
the three problems listed above.
Ovum
November 2003
75
The issue of whether mobile call termination prices should be cost oriented is
currently the subject of administrative and judicial review in a number of member
states. So we do not believe that it is appropriate to take a position in relation to this
issue. But we do recommend that the European Commission should:
• consider whether it should issue guidance requiring NRAs to ensure that mobile
operators charge regulated call termination prices in a non-discriminatory fashion
and, in particular, do not discriminate between charges to rivals and charges to
their own retail business
• consider what price differences between mobile operators should be allowed
when setting a regulated call termination price. For example, should differences in
call volumes, which lead to higher unit costs for smaller operators, be reflected in
the price differences? There are strong arguments that they should so as to
enable small 2G operators and greenfield 3G operators to compete with the large
incumbent mobile operators on more equal terms. We discuss this issue further
under Measure 7 of Chapter 8
• consider to what extent the arguments which lead to regulated prices for 2G
mobile voice termination should apply to other services such as SMS call
termination, 3G voice call termination and MMS call termination.
3.18 The impact of data protection on m-services
There are a number of elements of the Data Protection Directive63 that have the
potential to create a competitive disadvantage for new entrant providers of ECS. The
impact of this Directive is felt across the ECS industry but it has particular relevance
in the mobile sector. Accordingly, we assess this potential impact here.
Article 13(2) permits entities that obtain electronic contact details (for electronic mail)
for their customers in the context of a sale of a product or a service to use these
electronic contact details for direct marketing of their own similar products or services,
subject to the customers’rights to object (i.e., customers must ‘opt out’). Recital 41
sets out some of the policy background, noting that it was considered to be
reasonable, within the context of an existing customer relationship, to allow the use of
electronic contact details for the offering of similar product or services by the same
company. New entrants responding to our study have characterized the right of
existing operators to use contact data acquired from such past transactions to market
new services as skewing the playing field in favour of existing operators to the
acquisition of customers for new services. However, it does not appear to us that the
ability to use such contract details alone, not matched with other information
concerning matters including service usage patterns, for example, constitutes a
63
Directive 2002/58/EC of the European Parliament and of the Council of 12 July 2002
concerning the processing of personal data and the protection of privacy in the electronic
communications sector.
Ovum
November 2003
76
competitive barrier. Such contact details can be obtained elsewhere, relatively
cheaply, by new entrants.
Article 6(3) permits providers of publicly available ECSs to process traffic data to the
extent and for the duration necessary for the provision of value added services (i.e.,
any service which requires the processing of traffic or other location data beyond
what it necessary for the transmission of a communication or the billing thereof).
Similarly, Article 9(1) essentially provides that location data other than traffic data can
only be processed for use in the provision of value added services if the ECS provider
seeks the express consent of the user or subscriber to the transmission of such data
to the third party value added service provider. Recital 32 makes it clear that where
traffic or other location data must be forwarded from an ECS provider to the value
added service provider for the provision of a value added service, the subscriber or
user must be ‘fully informed’of this forwarding before giving their consent to the
processing of such data. The practical difficulties of obtaining such informed consent
where the user or subscriber is using a mobile handset are perceived as operating as
a practical barrier for independent value added service providers. The industry is
currently attempting to develop schemes to address such practical issues. However,
to the extent that such practical difficulties persist, it would appear that access to such
processed data by value added service providers that are not affiliated to ECS
providers is more difficult than for ECS providers self-supplying value added services.
The Study Team understands that the industry is currently working to produce
practical solutions to address the potential for competitive disadvantage that Article
6(3) and 9(1) might represent for mobile value added service providers that are not
affiliated to ECS providers processing traffic data (and location data other than traffic
data). Accordingly, it makes no recommendations.
3.19 Mobile content
A number of mobile operators have entered into arrangements to acquire the
exclusive 3G rights to premium content. For example, T-Mobile has the rights to the
Bundesliga and H3G has the right to the FA Premier League in the UK, Seria A in
Italy and ice hockey in Sweden. Respondents to our study have indicated that they
believe that sports rights are important in driving early adoption for 3G. All of our
respondents have indicated that, at the moment, it appears that such rights are being
acquired through ‘highest bidder wins’contests. There appears to be no evidence
that rights holders are currently placing weight on the size of the bidding mobile
operators’subscriber bases (i.e., ‘eyeballs’controlled). As such, it does not appear
that ‘incumbency’currently provides an advantage in this respect. However a number
of respondents believe that such a development might occur in future. At the same
time the mobile operators with the largest subscriber base or those with the greatest
financial strength are also in a position to make higher bids.
The respondents to our study noted that the content for 3G market is so embryonic
that it is virtually impossible to anticipate how it will develop during the life of the
current arrangements. No mobile operator has any certainty that the exclusive rights
that it has acquired will actually make any difference at all to take up or ARPU,
Ovum
November 2003
77
irrespective of the amount of market research that it may conduct. Our respondents
have noted that they are expecting quite different dynamics for non-sport content. In
particular, they expect that games will be very fragmented, and might fragment across
the competing platforms in a manner that reflects the target demographics of each
operator (leaving aside the technical issues associated with multiple game
compatibility for the operating systems).
One respondent has indicated that it believes that 3G data services will not be ‘mobile
Internet’. It expects that web browsing will only account for a relatively small
proportion of traffic (when compared to anticipated peer-to-peer messaging traffic
flows, for example). However, another uses an Internet content analogy to describe
mobile data services (to argue that rights to mobile content are not analogous to
broadcasting rights). It would appear that the views in this respect reflect the varying
business plans and target demographics.
While it is too early to identify mobile-specific issues relating to premium content, the
Commission should monitor rights acquisition, to ensure that such arrangements do
not become barriers to competition. The Study Team’s discussion of potential
barriers constituted by rights acquisition and exercise in the broadcasting context is
also relevant in the mobile context.
Ovum
November 2003
4
78
Broadcasting services
This chapter reviews competition in broadcasting-related markets of the European
Union to identify potential barriers to competition. Our discussions with industry
indicate that there are concerns relating to four key potential barriers to competition:
• exclusive rights to content, both in terms of the potential foreclosing effect of such
rights on new platform operators when such rights are held by existing operators
and in terms of the relationship between rights holders and platform operators,
where a particular rights holder is not vertically-integrated with the particular
platform operator;
• the potential for the scope and effect of the existing regulatory structures for
different types of services to skew the basis on which different but competing
services are able to compete;
• the regulation of CAS, and the need to ensure that appropriate and proportionate
regulation is imposed, rather than maintained on the basis of a presumption of
proportionality and appropriateness; and
• the need to ensure that regulatory measures intended to facilitate the
development of, and access to, iTV and other e-Europe services are directed at
doing so.
In order to better understand the origins of these concerns, we first briefly describe
the broadcasting value chain, and follow it with a discussion which focuses on these
concerns, as illustrated by the following individual sets of issues:
• exclusive rights to content and the balancing of the rights of rights holders and
platform operators;
• the impact of “must carry” obligations on the relationship between program
providers and platform operators;
• the changes in the types of content that can be provided, whether driven by the
‘converging’environment or other factors, and the need to ensure that regulation
does not delay or skew the development of such changes;
• the implications of variation in the number of competitive platforms;
• the role, and appropriate regulation, of conditional access systems; and
• the need for thorough analysis and assessment of perceived competitive
problems that relate to consumer access to interoperable iTV services, to
portable content authoring and to access to APIs, to ensure that regulatory
measures are proportionate and appropriate.
Ovum
November 2003
79
4.1 The broadcasting services value chain
The broadcasting sector is characterised by multiple, complex relationships between
entities at different levels of the value chain. There are essentially three general
layers or categories, which depend on the nature of the relationships between the
entities in the broadcasting value chain: (a) a wholesale market for ‘rights’; (b)
wholesale markets for access to network and/or platform infrastructure (including
‘technical services’); and (c) a retail ‘broadcasting services’market. Additional
elements can be identified in a number of different broadcasting services business
models, including advertising and the potential need for a ‘return-path’(particularly in
an interactive environment). The value chain is illustrated in Figure 4.1 below:
Figure 4.1 The value chain for TV broadcast services
Wholesale Provision of Content
At the top of the value chain, there is at least one broad market for the production and
supply of content for television. Within this broader market, there are a number of
Ovum
November 2003
80
different ‘types’of content that are provided as television programming in most
Member States,64 namely:
• ‘premium’films
• ‘premium’sports
• non-premium programming for television
• interactive programmes or features, often complimentary and linked to other
programming.
This content is turned into ‘channels’by channel providers, through the acquisition
and packaging of programming to appeal to consumers.65 Channel providers
purchase programmes either from independent producers or produce them internally
(or through related entities).66 Most films are licensed from the major studios, but
some are commissioned or produced by channel providers. Sports rights are
generally licensed from the relevant rights holder. The dynamics of acquiring licences
to rights for sports that are ‘listed’under the measures implementing the TWF
Directive differ, since no channel provider can acquire exclusive live rights to such
events.
Content tends to be exclusive to a single channel in each geographic market (and is
protected by the terms of the licence to each channel provider). The rights for sport
and films are often referable to a ‘window’for broadcast (e.g., rights to broadcast
during a match, highlights immediately following the match or the right to broadcast a
film immediately following its theatre run). As a result of the exclusivity associated
with most programming, channels are strongly differentiated (i.e., no two channels
carry the same programming). In turn, channel providers aim to differentiate their
channels on the basis that the more attractive the channel, the greater the price that
can be charged (at both the wholesale and retail levels), reducing the ability of other
channels to constrain pricing behaviour.
64
In certain circumstances, each might constitute ‘markets’or ‘sub-markets’in their
own right.
65
In the context of advertiser-supported TV, it also typically involves the insertion of
advertising (for which the channel supplier earns advertising revenue). Rarely, the
channel supplier may leave ‘availability’for advertising to be inserted by distributors.
66
In this respect, it appears that it is audience demand, more than the quotas in the
Television Without Frontiers Directive (the “TWF Directive”) that drives local content.
There is currently a debate as to the role and purpose of such quotas and possible
unintended effects on the market (e.g., some suggest that they operate as barriers to
entry for new entrant broadcasters unable to fund the more expensive “productions”)
and their future role in an increasingly multi-channel specialist environment.
Ovum
November 2003
81
Relationships Between Channel Providers and Platform Operators
Channel providers either distribute channels on their own account67 or supply them to
network or platform operators for distribution. The relationships between channel
providers and network or platform operators can be complex. Increasingly, network
and platform operators control or have a stake in at least one channel provider. In the
analogue distribution environment this has led to some concerns about discrimination
in favour of, or undue preference shown to, channel providers related to the platform
or network operators. In the digital environment, there is, inherently, greater capacity.
However, a number of services that are already being and will in the future be
provided on digital networks require greater capacity. It is not yet clear how these two
factors will impact on the competing commercial incentives of network or platform
operators: namely, to carry as much high quality content as possible and to ensure
capacity for innovative services.
The second, related, factor that can affect such relationships is the contractual terms
on which channels controlled by a vertically integrated channel provider and
distributor are provided to competing distributors. There have been concerns,
expressed in a number of Member States, about the extent to which such vertically
integrated entities use deep discounting and other discounting practices that could
produce a price squeeze.
The third relevant factor is the ‘must carry’obligation. This obligation is imposed on
entities providing networks used for the distribution of radio or television broadcasts to
the public where a significant number of the end-users of the networks use it as their
principal means to receive such broadcasts. It essentially requires such entities to
carry designated channels and services. The new Universal Services Directive
permits Member States to determine appropriate remuneration for the carriage of
such channels and services.
Retail Supply
Free-to-air distribution is available to viewers with the appropriate reception
equipment, without subscription.68 Historically, free-to-air distribution has been
analogue. However, the number of digital distribution platforms is slowly growing.
PayTV distribution supplies channels to viewers who, through a subscription contract,
are authorised and have the appropriate equipment to receive them. It typically
entails the retailing to subscribers of offerings (or ‘bundles’) of ‘basic’and ‘premium’
channels. In most Member States, bundles of ‘basic’channels are offered, consisting
67
e.g. the analogue free-to-air model or other model based on vertical integration
with a platform or network operator or some pay TV channel providers that are not
vertically-integrated, notably Canal+ outside France.
68
Some Member States have published ‘target dates’for analogue switch-off,
predicated on a specified level of digital penetration (e.g., between 2006 and 2010, if
95% digital penetration).
Ovum
November 2003
82
of between twenty and sixty-five basic channels. These offerings are cheaper than
premium channels. Premium channels are also available, often bundled, and often
‘themed’. In a number of Member States, the pricing of such premium channels
varies by reference to the number of premium channels taken by a subscriber.
The retail distribution markets appear to operate differently across the Member
States. For example:
• the French market has been characterised as a single market including both
digital terrestrial and payTV, whether by cable or satellite;
• the German market has historically been identified as having separate terrestrial
and cable markets;
• the United Kingdom has been characterised as having a digital broadcasting
sector in which terrestrial TV and other forms of audio-visual entertainment are
seen as close substitutes for television distributed over cable and satellite
platforms; and
• Italy has been found to have separate markets for the provision of digital
terrestrial services and digital satellite services.
By and large, these differences appear to reflect the extent of inter-platform
competition and the respective timing and maturity of the introduction and
development of each platform in each geographic market.
The pervasive trend with digital programming is to incorporate ‘interactive’elements
into programmes. There are currently two basic types of interactive services: digital
data services that include services that are essentially digital Teletext (e.g., walled
gardens containing text based on web pages that has been reauthored for TV
screens), transaction services (e.g., pizza delivery), games and voting (including
‘participate at home’functionality for game shows or competitions). In addition, there
are enhanced TV services, in which clicking on icons takes you to a linked page or
video and/ or audio stream (e.g., ‘change the camera angle’functionality for sporting
events, ‘select the story’functionality for news). Currently, most network or platform
operators provide such functionality using additional transmission channels and a
return path (the latter may run over another network or platform). There is
increasingly heavy cross-promotion of enhanced TV services by its providers, both on
channels that have the embedded links and on other channels that they supply. As
such, it may be necessary for the Commission to revisit whether retail broadcasting
services and such interactive services continue to fall into different markets.69
69
See, for example, paragraph 4.1.1.3 in BiB Case No IV/36.539, 98/C 322/05 and
paragraph 30 in BSkyB/ Kirch Pay TV Case No COMP/JV.37.
Ovum
November 2003
83
4.2 Potential Barriers to Competition From Rights
Acquisition and Exercise
As noted above, the link between content, particularly premium content, and the ECS
relevant market for broadcasting transmission services will become increasingly
important as Member States implement the new EU framework. A number of the
respondents to our study have raised concerns about the potential foreclosing effect
of the control of rights to premium content. For example, concerns were expressed
about platform operators with agreements conferring rights to a large volume and
wide variety of such rights and the advantages held by some platform operators as a
result of their greater scale (i.e., in any ‘bidding’environment or “bidding market”
scenario).
In relation to non-broadcasting rights to premium content (e.g., UMTS and Internet
rights), a number of respondents to our study highlighted the scale advantages
conferred on entities with large existing customer bases (e.g., large numbers of
‘eyeballs’) in acquiring rights, both in terms of the ability to pay for content and as a
result of the desire of holders of certain types of content to maximise the number of
‘eyeballs’to which they have access. However, other respondents noted that it
appears to them that the acquisition of such rights is highly competitive, as competing
platform operators attempt to acquire content that they believe will help stimulate
platform migration by early adopters. They also note that most of such negotiations
are with rights holders that are in extremely powerful negotiating positions, and that
most such contracts have a relatively short term.
Rights to Premium Content
Movie Rights
There have been a number of instances in which there have been concerns that
payTV operators (initially Kirch) have, through exclusive licensing arrangements with
Hollywood studios (or distributors), accumulated ‘libraries’of premium films that give
them an unassailable position in relation to the onward supply of such films.70 For
example, Canal+ has been described as enjoying a “de facto monopoly position on
premium films for payTV”in various geographic markets.71 Concerns that the merger
would create or strengthen a dominant position in the market for the acquisition of
exclusive rights for premium films were expressed by the Commission in relation to
the recent merger of Spain’s Via Digital and Sogecable, as a result of the
Commission’s analysis of that case before its referral to the Spanish authorities. The
Spanish Competition Court’s opinion in relation to this merger considered that the
merger should be subject to a range of conditions, including one restricting the term
70
BSkyB/ Kirch Pay TV Case No COMP/JV.37 at para. 47 ff and Bertelsmann/ CLT
Case No IV/M.779 at para. 33.
71
Vivendi/ Canal+/Seagram Case No IV/M.2050 at para. 40.
Ovum
November 2003
84
for which it acquires first and second window television broadcasting rights from
“major studios” to a maximum period of one year. It remains to be seen how matters
like rights to premium or “must have”content affect assessments of the market power
of vertically-integrated platform operators/broadcasters in the context of analyses of
the provision of wholesale broadcasting transmission services under the new ECNS
framework.
In January 2003, the Commission’s DG Competition opened an investigation into the
sale of TV rights to films by Hollywood studios to European payTV operators. The
Commission is concerned that the current contracts stifle competition between
studios and that long-term exclusive agreements (particularly where one payTV
operator is a party to many such agreements) prevent entry into the payTV market.
The inquiry is considering clauses dealing with duration (generally five years),
exclusivity and payment terms, including terms requiring payTV operators to match
payments made to one studio to those paid to its rivals (so called “most favoured
nation” clauses), escalator payments of 3 per cent per annum, “blockbuster” clauses
(which set payTV prices by reference to box office receipts) and additional fees tied to
subscriber growth. The Commission has indicated that it is concerned that these
clauses have the effect of facilitating collusion between studios in setting prices and
terms. The investigation follows a complaint by the French satellite platform
broadcaster TPS about arrangements between several studios and Canal Satellite.
Canal Plus has filed a separate complaint in relation to TPS’arrangements with TF1
and M6.
Sports Rights
In June 2002, DG Competition closed its investigation into the sale of TV rights to the
Champions’League, after UEFA revised its policy for selling media rights to the
Champions League. The new UEFA rules permit the clubs to sell their matches
individually if UEFA has not sold the rights to their Tuesday night matches; ensure
that all rights will be offered to the market (including Internet and UMTS rights); permit
the exploitation of deferred TV rights; entail splitting the media rights into 14 smaller
packages (some of which are co-exploited by UEFA and individual clubs); and limit
the term for which rights contracts will be awarded to three years.
It also appears that the Commission’s current FA Premier League investigation is
considering the impact of the control over Premier League media rights exercised by
the FAPL, the extent of rights retained by the clubs, the number of games currently
broadcast, the effect of the term of the exclusive contracts for rights and the potential
for accumulation of a large number of such exclusive, long-term rights to foreclose
market entry by other entities. It is anticipated that the investigation will be concluded
this year.
Interactive and other related Content
A number of respondents to our study noted issues associated with so called
‘interactive’elements in programming. In particular, they referred to enhanced TV
services (with embedded icons that allow end-users to ‘click through’to a linked page
or video and/ or audio stream) and branding. Interactive services are usually heavily
Ovum
November 2003
85
promoted both on the particular channel with the embedded link and on other
channels provided by the rights holder. In a number of Member States (e.g., Italy and
the United Kingdom) there are rights holders that are required to provide their
programmes or channels to other platform operators. However, they are not required:
• to provide linked interactive services
• to remove the links to such services
• to provide so-called “clean feed”, allowing the replacement of such services with
other interactive services developed or provided by the competing platform
operator.
Anecdotal evidence suggests that these practices have a significant impact on enduser perception of the competing platform operator and on the ability of competing
platform operators to develop and provide their own enhanced TV services (even
though such competing operators use infrastructure that is better equipped to provide
return-path based services).
Based on this analysis we recommend that:
• close attention is paid in future to exclusive rights to broadcasting content,
particularly where such rights are of broad scope (both in terms of platforms and
number and variety of rights) and of long duration;
• competition law processes are used for such scrutiny;
• the European Commission considers the development of guidelines based on
competition case law to date (as described above); and
• the issues identified above associated with embedded interactive content be the
subject of further study. The extent to which such services are promoted with and
within conventional broadcasting services is such that the approach taken in BiB
and similar cases, considering interactive services and broadcasting services
separately, may warrant reconsideration.
Channel Supply and Access to Platforms/ Networks
As noted above, the relationships between channel providers and network and/or
platform operators are complex. In essence, they are the commercial point at which
the channel providers’view that platform operators provide them with transmission
services and the platform operators’view that channel providers provide them with
content intersect most directly.
The relationship has been characterised by the Commission, in a series of merger
reviews, as the provision of broadcasting transmission services to channel
suppliers. It has been considered in assessments of the extent to which the different
‘means of transmission’are substitutable to meet the requirements of broadcasters
(channel suppliers). There has been variation in relation to such substitutability
between cases relating to different Member States. For example, in the early
German case MSG Media it was found that cable and satellite transmission or
Ovum
November 2003
86
transponder capacity were not substitutes.72 Similarly, in Nordic Satellite Distribution,
transponder capacity was found to be in a stand-alone market. These cases should
be contrasted with the approaches taken in relation to the United Kingdom (i.e.,
Sky/BSB and BiB/Open) and France (i.e., TPS). It should also be noted that, in the
more recent German cable mergers (Blackstone/ CDPQ/ Kabel Nordrhein-Westfalen
and Blackstone/ CDPQ/ Kabel Baden-Wuerttemberg), it was acknowledged that the
rapid changes occurring in the competitive environment might mean that the market
for transmission capacity might now encompass an aggregate of all available
platforms.73
The characterisation of the relationship as the provision of broadcasting transmission
services is also reflected in the new electronic communications framework, through
the inclusion of a relevant market defined in the Commission’s Relevant Markets
Recommendation74 as ‘broadcasting transmission services and distribution networks
in so far as they provide the means to deliver broadcast content to end users’.
However, the impact of this approach, when combined with the treatment of
‘broadcasting platforms’as ECNs, under the new ECNS framework is unclear. For
example, the range of obligations that might be imposed on any provider of wholesale
broadcasting transmission services found to have SMP includes an “access”
obligation. It is far from clear what “access” will mean in this context. Content
providers are expressly not providers of ECSs and are, therefore, on the face of the
regime, outside its scope. However, the conferral of ‘access’rights on content
providers could be seen as essentially conferring on them at least some of the
benefits of the regime (without the attendant obligations). While it seems unlikely that
the policy imperative for conferring access rights on providers of ECSs currently apply
to content providers, it is not clear what form any “access” obligation might take in this
context. It would appear to be important for the Commission and the NRAs to
consider this matter, both to avoid the implementation of the ECNS regime with
unintended effects and to ensure that NRAs are able to use the access remedy (when
it is appropriate and proportionate to do so).
72
It is, in this context, important to note the effect of the particular legal and
regulatory conditions that existed (e.g., the control by building owners over the final
cable connections to viewers and the restrictions on installation of multiple satellite
receivers on multi-dwelling residential properties).
73
For example, construction of multiple networks providing the final cable
connections to viewers. See, further, Communication from the Commission to the
council, the European Parliament, the European Economic and Social Committee and
the Committee of the Regions on the transition from analogue to digital broadcasting
(from digital ‘switchover’to analogue ‘switch-off’SEC(2003)992.
74
Commission Recommendation of 11/02/2003 on Relevant Product and Service
Markets within the electronic communications sector susceptible to ex ante regulation
in accordance with Directive 2002/21/EC of the European Parliament and of the
Council on a common regulatory framework for electronic communications networks
and services.
Ovum
November 2003
87
The alternative characterisation of the relationship is to view the content as the
‘asset’which the platform or network provider seeks the right to use. In view of
the current position under Community law, it appears that it would be difficult to base
an access request on an ‘essential facilities’case, to compel a rights holder to license
such rights. While it may not be the case that a platform provider would have to
demonstrate that it requires access to provide a “new product”, the platform provider
would have to show that the rights holder occupies a dominant position, that the rights
holder possesses a facility that is essential for the supply of certain services, that the
rights holder has refused to grant non-discriminatory access (without an objectively
justifiable reason) and that the rights are acquired in an upstream market as a
necessary input for supply in a downstream market.75 Clearly, this would not be an
easy case to sustain.
However, we note that the Commission’s current investigation into the agreements
between Hollywood studios and payTV operators may raise issues about the rights
(particularly premium content rights) held by payTV operators in the context of their
access to transmission services required for retail distribution. In particular, there
may be significant issues relating to foreclosure, non-discriminatory access and
transparency. These issues are most starkly relevant in a geographic market in
which at least one payTV rights holder is vertically-integrated with a distribution
platform. For example, the Sky ratecards were given by Sky as non-statutory
undertakings, in 1996, to meet concerns of the DG OFT about competition in the
market for wholesale payTV. More recently, the OFT considered potential breaches
of Chapters I and II of the Competition Act by Sky (finding dominance on particular
markets but not the alleged margin squeeze).
It appears to the Study Team that the ECNS regulatory framework, because it applies
to “broadcasting transmission services”, raises the very real possibility that content
providers will be able to acquire ‘access’to transmission services in the same way
that providers of ECSs may acquire access. We, therefore, recommend that the
Commission takes the necessary steps to ensure that the balance between the
competing rights of ECNS providers and rights holders is maintained.
The Impact of the Number of Platforms
There is real doubt in a number of the smaller Member States that DTT platforms will
be deployed. In such circumstances, there is a risk that some Member States may
have a single digital network or platform that is capable of distributing digital
broadcasting content. Clearly, if a Member State were to have a single digital
platform (without multiple providers using such a platform), regulators will need to
consider the access obligations that may be warranted or appropriate (e.g., in view of
75
The final element might be found to be unnecessary in the IMS case, which is
currently before the Court of First Instance.
Ovum
November 2003
88
the availability of spectrum, technical characteristics of the network) in such a
situation.76
In a number of Member States, the commercial provision of ADSL-based delivery of
broadcast (or narrowcast) content is being considered, and a number of pilot
schemes have been successfully tested. However, in recent transactions, concerns
have been expressed about the potential loss of the competitive opportunity
represented by such an alternative platform. For example, in the Via Digital and
Sogecable merger, the Commission was concerned that the interest that Telefónica
would acquire in the merged entity would affect its development of payTV over ADSL
through project Imagenio. The entry of such new platforms may offset the absence of
DTT networks in some Member States, as long as they are not controlled by entities
that already have an interest in existing platforms (e.g., fixed incumbents with an
interest in the cable or satellite platform).77
It is also worth noting the consolidation that is beginning to happen across the
Member States, essentially removing or reducing competition between multiple
platforms using the same technology (e.g., Telepui and Stream in Italy and Sogecable
and Via Digital in Spain).
The Study Team recommends that the Commission monitor the development of
digital platforms to ensure that it is aware of the number and capability of such
platforms in each Member State, and considers the implications of reductions in the
number, and concentration in relation to the ownership, of such platforms.
Short term gains from allowing investors with the ability to invest in multiple platforms
to do so might be overborne in the medium to long term by market parallelism
designed to erect entry barriers, to maintain margins through exploitation of
externalities or to ensure that new more efficient services that might cannibalise
revenues from inefficient services do not develop. It is imperative to ensure that the
implications for infrastructure-based competition are carefully considered in the
context of forward-looking reviews conducted in relation to the approval of
transactions that permit entities to acquire interests in multiple platforms, and that the
regulatory trade-offs introduced to counteract the loss of alternative platforms are
carefully considered.
76
For example, Directive 95/47 was drafted to reflect the possibility that only a single
pay TV platform might develop in each Member State. As such, it imposed the “fair,
reasonable and non-discriminatory access” obligation on all CAS providers, rather
than those with a degree of market power.
77
In fact, there is a concern that sustainable DTT entry may not occur in some
Member States in which there have been strong analogue pay TV platforms.
Ovum
November 2003
89
4.3 Consistent Regulatory Treatment of Competing
Services
Must Carry Obligations
The nature and scope of must carry obligations vary across the Member States. The
inclusion of Article 31 of the new Universal Service Directive might go some way
towards harmonising the approaches taken. However, it is unlikely to lead to the
review of the measures empowering media authorities to approve changes to
bouquet line-ups and pricing that currently exist in a number of Member States. It
might also skew the commercial arrangements between channel providers providing
channels with must carry status and platform operators. It should also be noted that
the growth of “must offer” or “must have” content, and the role of such content in
multi-channel environment is not addressed by Article 31. Such content could
become increasingly more significant than “must carry”. In addition, there are a
number of features of Article 31 that should be noted, namely:
• the obligation may be imposed on undertakings under a Member State’s
jurisdiction providing ECNS used for the distribution of broadcasts - this
formulation does not address the difficulties that have already arisen in relation to
jurisdiction over undertakings providing broadcasting channels and services over
satellite platforms;
• the ‘broadcast channels’and ‘services’concepts are not defined in the Universal
Service Directive. Given that the terms do not reflect the terms used in either the
TWF or Information Society Directives or elsewhere in the ECNS Directives, there
appears to be the potential for differing interpretations (and implementation)
across the Member States;
• the potential scope for the imposition of must carry obligations on networks used
as the principal means by which a significant number of end-users receive radio
and television broadcasts is broad. While the inclusion of this concept was
intended to be forward-looking, technology-neutral and to protect cultural
interests, when combined with the undefined ‘broadcast’concept it may have the
unintended effect of providing scope for the extension of must carry rights across
a broad range of content and multiple platforms in a manner that stretches the
public policy justifications for such obligations. As a practical matter, it may
become increasingly difficult to gather the empirical data necessary to determine
whether ‘a significant number’of end-users use one means or another for their
reception, or to identify which is the ‘principal’means used when end-users take
advantage of multiple means;
• there appears to be a disconnect between the types of network providers which
must be considered in the context of determining non-discriminatory remuneration
(e.g., potentially all undertakings providing ECNs) and the undertakings which
may be obliged to carry must carry channels (e.g., operators of ECNs relied on by
a significant number of end-users as their principal means of receiving
broadcasts); and
Ovum
November 2003
90
• it does not appear that Article 31 will impact on the powers of media authorities
(including local authorities) in a number of Member States to improve the channel
line up, and pricing, of ‘basic’service packages. As such, the potential for
increasingly diverging powers to control bouquets offered is preserved.78
Regulators might also consider whether must carry-type obligations in their current
form will continue to be appropriate in the future, in a ‘digital’multi-platform
environment, where much of the content currently benefiting from must carry status
might become ‘must have’content for platform operators.79 In this context, we note,
for example, that subscriber numbers for Sky’s Irish operations grew significantly
when it started to carry BBC1 and BBC2 and, subsequently, RTE. In a digital
environment, with the potential for the supply of a greater volume and range of
content, there will be strong commercial incentives for platform operators to carry high
quality content, particularly where rights to such content are exclusive. It is these and
similar dynamics that drive the imposition of “must offer” requirements.
Industry remains concerned about the potential for ambiguity and inconsistent
implementation across Member States in relation to the applicability, scope, nature
and form of must carry obligations to skew inter-platform competition, and to
perpetuate a distinction between entities that ‘have’and entities that do not ‘have’
must carry status. That distinction becomes increasingly difficult to justify in the
digital environment.
The Study Team recommends that the Commission considers whether the output
from its current study relating to must carry issues addresses the issues raised
above.
Varying Regimes for Different Services
A number of respondents to our study have raised concerns as to the
appropriateness of applying differing specific regulatory regimes to services in the
converging digital environment, and whether this will be appropriate, sustainable or
reflective of the reality of the market. From 25 July 2003, the following types of
services will be regulated in different ways under distinct regulatory regimes:
• ECSs – a service normally provided for remuneration which consists wholly or
mainly in the conveyance of signals on ECNs, including transmission services in
networks used for broadcasting but expressly excluding services providing or
exercising editorial control over content transmitted using ECSs, and which also
78
It is clear, of course, that the Commission’s position that provision of retail broadcasting
services to end-users is not an ECS further dilutes any incentive (and, perhaps, remit) to
consider whether such national measures amount to must carry measures.
79
It is clear, however, that some must carry content, with lower viewer numbers and,
accordingly higher opportunity costs, is unlikely to benefit for the commercial need for
‘attractive’must carry content (e.g., the UK Parliament channel).
Ovum
November 2003
91
does not include information society services which do not consist wholly or
mainly in the conveyance of signals on ECNs;
• information society services – any service normally provided for remuneration, at
a distance, by electronic means and at the individual request of a recipient of
services, and which includes services like video-on-demand; and
• television broadcasting – initial transmission of television programmes intended
for reception by the public, including the communication of programmes between
undertakings with a view to their being relayed to the public. It does not include
communication services providing items of information or other messages on
individual demand such as telecopying, electronic data banks and other similar
activities.
The gaps and overlaps between, and ambiguities in relation to the boundaries
between, these concepts and the significantly different impact of the application of
one regime or another, is already beginning to open up the potential for regulatory
lacunae. For example, the Commission has taken the view that the ECS concept
does not include the delivery of retail ‘broadcasting’services or end-user access to
networks for the receipt of such services. The impact of this interpretation is to
empower regulators to monitor consumer protection issues and elements of
consumer contracts in relation to telephony and Internet services provided over cable
networks, but arguably not in relation to broadcasting services provided over the
same platform. This may reflect the fact that telecommunications regulation has
historically recognised the concept of “end-user access to networks”, while
broadcasting regulation does not, and illustrates the inherent conflict between the
historic approaches to communications infrastructure and (broadcasting) content
regulation.
In addition, when read with the TWF Directive, the exclusion of retail broadcasting
services from the ECS regime (i.e., leaving it to the Member States to regulate such
services) may have the effect of continuing to permit Member States to retain existing
unharmonised and, in some cases, heavy regulation of retail broadcasting over
terrestrial and cable platforms, while applying the country of origin principle to satellite
platform-based retail broadcasting. As a result, terrestrial and cable broadcasting
platforms may continue to bear the costs (monetary and otherwise) of such
regulation. As such, it may have a skewing effect on the market, inhibiting the ability
of terrestrial and cable platform operators to compete effectively.
In addition, the definitions may mean that ‘modularised’80 content is regulated
differently, depending on the particular digital modules put together to form a service.
The form and structure of content is evolving in the digital environment. It is
80
Increasingly, content is being produced as digital ‘modules’that can be put
together for different platforms, as required (e.g., the audio and video modules for
television, perhaps with the addition of text and interactive modules, the audio module
for radio, an enhanced text module with the audio and video and interactive modules
for the Internet).
Ovum
November 2003
92
anticipated that this trend will continue, with content increasingly being produced in
this manner. At the same time, content is increasingly being delivered over platforms
other than those with which it is traditionally associated. It is, and will increasingly be,
provided over alternative, competing or complementary, platforms.
The study team recommends that the Commission undertakes a comprehensive
review of the overlaps of, gaps between and boundaries of, these regulatory
distinctions
Given the digital economy aims of the Community, such a review appears necessary
to ensure that the application of overlapping and possibly inconsistent regulation does
not defeat such aims.
4.4 Access to Conditional Access Systems
Conditional access systems (“CAS”) are essentially technical systems that ensure
that only viewers authorised to acquire a service (whether as part of a subscription or
on demand) are able to do so, including authorisation signals (i.e., entitlement control
messages, entitlement management messages and verification) in the data stream
and provide message processing services (e.g., decryption and unscrambling). They
also prevent unauthorised signals from corrupting the decoder population. The
conditional access system is integrated in the set-top box (“STB”), as illustrated in
Figure 4.2 below.
All operators of conditional access services who provide access services to digital
television and radio services and whose access services are depended on by
broadcasters to reach any group of potential viewers or listeners are currently (and
will continue to be) required to offer to all broadcasters, on a fair, reasonable and nondiscriminatory basis, technical services enabling such broadcasters’digitallytransmitted services to be received by those authorised by means of decoders
administered by the service operators. Article 6 of the Access Directive permits the
review and revision of these obligations in relation to operators that do not have SMP
and to amend the obligations imposed on those entities with SMP, through a market
analysis conducted under Article 16 of the Framework Directive.
Ovum
November 2003
93
Figure 4.2 The architecture of STBs, showing CAS and API
There are two issues that flow from the regime (with the potential for review)
described above:
• what does the ‘fair, reasonable and non-discriminatory’obligation actually entail;
and
• should this ‘remedy’be revised, as permitted by the new electronic
communications regulatory regime.
With the exception of relatively recent review in the United Kingdom, prompted by the
terms of access to SSSL’s CAS, there has been little regulatory attention to
interpreting the fair, reasonable and non-discriminatory test. Oftel has taken the view
that such terms are those which would be most likely to maximise efficient use of the
CAS, maximising entry to the platform and choice to the consumer, and reflecting the
willingness of a rational customer to pay for the service. It has gone on to take the
view that a conditional access service provider would be ‘expected to recover their
costs and make reasonable, but not excessive profits’. Finally, Oftel has taken the
view that different terms can be offered to different customers (reflecting the different
values that they might place on the service), only if such differentiation does not have
a significant adverse effect on competition. In its view, permitting new players to join
the platform ‘at a discount’when it would have been uneconomic for them to join on
more typical terms will be pro-competitive.
The new electronic communications framework permits the review and amendment or
withdrawal of the “fair, reasonable and non-discriminatory terms” obligation. As such,
NRAs may be permitted to consider whether this remains the appropriate ‘remedy’.
In considering this issue, they might wish to consider whether:
Ovum
November 2003
94
• an access obligation in relation to CAS is actually the appropriate manner in
which to address the perceived regulatory mischief (i.e., whether it is really
access to the CAS, rather than, for example, APIs and authorising tools, that is
the regulatory concern);
• they share Oftel’s view in relation to the nature and effect of the current
obligation;
• other access obligations in the Access Directive would be more appropriate
(bearing in mind that such obligations can only be imposed on entities with
significant market power in the provision of CAS (i.e., there is no power under the
new regime to redefine the scope of the CAS ‘market’));
• the obligation in relation to CAS might become more ‘real’for those systems
which are currently closed, if they are subjected to access obligations relating to
their transmission services under the new regime (e.g., cable systems);
• the debate surrounding the extension of this remedy to “must carry”that arose in
the course of the adoption of the new ECNS regime; and
• in a ‘converging’environment, the benefit of access should only extend to
broadcasters, or whether other providers of interactive services should benefit
from the same or another form of access.
It should also be remembered that the standardisation powers set out in Article 17 of
the Framework Directive provide a mechanism for identifying standards and/or
specifications to serve as a basis for encouraging the harmonised provision of
associated facilities (which include CAS), where the perceived regulatory problem
relates to standards and/or specifications.
It is important that regulators consider the likely impact of: (i) lifting the fair,
reasonable and non-discriminatory access obligation off operators without SMP; and
(ii) changing the nature of the access obligation itself.
4.5 Interoperable iTV services and APIs
A number of respondents to our study have indicated that industry remains concerned
that the existing regulatory framework applied to CAS may be extended to APIs
without a thorough analysis and assessment of perceived competitive problems that
relate to ensuring interoperability of iTV services.
Accordingly, the Commission has asked us to consider whether the development of
common authoring tools can address perceived barriers to the development and
delivery of interactive television, or iTV, and whether this would reduce the need to
consider mandating API standards or imposing API access obligations. In particular,
we have been asked to consider the links between such authoring tools,
interoperability and access.
The European institutions have recently re-emphasized the importance that is placed
on developing Information Society services and other e-services, and ensuring that
they are available to as many European citizens, over the broadest range of platforms
Ovum
November 2003
95
and using the widest variety of end user terminals as is possible.81 In this light, this
chapter briefly describes the types and forms of iTV content to which that term refers,
before identifying the potential barrier to competition, otherwise characterised as an
interoperability issue, that we consider to be relevant to authoring tools (i.e.,
portability of content/ applications). In doing so, we will seek to draw out each factor
that impacts on cross-platform portability of iTV services. It concludes by assessing
the extent to which the development and adoption of common authoring formats
might facilitate porting.
What is iTV?
We have considered the types of services that should be encompassed in an
examination of the potential role of authoring tools in relation to the development and
delivery of iTV. In essence, this chapter focuses on interactive services that are
designed specifically for TV, both:
• enhanced TV services (based on the programmed or cyclical downloading of data
associated with signals from a linear TV service), including services providing
viewers with the ability to bet on the video content, choose camera angles or
‘play-at-home’quiz games. A return path may or may not be required;82 and
• non-TV-related or ‘interactive’services (e.g., banking, shopping and games).83
81
See, inter alia, the “Communication from the Commission to the European
Parliament, the Council, the European Economic and Social Committee and the
Committee of the Regions on Barriers to widespread access to new services and
applications of the information society through open platforms in digital television and
third generation mobile communications”, 9 July 2003, COM(2003)410 final.
82
Enhanced television consists of the transmission of the televised programme
along with related complementary information. This information can be any
combination of texts, images, sound and/ or video sequences, and is usually linked to
the course of the programme. As such, it is valid for a limited duration (defined by the
programme broadcaster). The video stream may not be interrupted by accessing
complementary information, in that information continues to be displayed as a
‘vignette’or in the background (where the complementary information is embedded in
the video image) or the viewer may have to quit the programme in order to access to
complementary information. Enhanced television relies on local interaction between
the user and the application temporarily residing in the receiver, rather than a return
channel.
83
Non-related TV services also involve the broadcast of textual information, sound,
images and video destined for TV reception, but does not necessarily imply the
presence of, or any links to, a continuous stream of video images. These
applications generally require an interaction channel, usually via integrated modem
communication (over the PSTN or a bi-directional cable network).
Ovum
November 2003
96
The chapter does not expressly consider VoD or NVoD, because such services
currently operate essentially as advanced pay-per-view services (allowing viewers to
choose both the programme and the viewing time). Nor does it consider Internet
access services accessed via TV sets, entailing “full screen” Internet browsing using
a keyboard and Internet-specific software. However, we note that such services must
be considered in developing medium-term iTV policy, to ensure that the service and
platform convergence that they represent is not hampered by policy decisions.
Finally, we note that the production and delivery of, and consumer demand for, iTV
content and applications is currently embryonic at best across the Member States. In
addition, it does not appear that there is a significant amount of pent up demand for
such services. A Statistical Research Institute survey performed in the second half of
2002 indicated that 72% of the consumers surveyed did not want iTV services, with
very similar results among those have currently have, and those who do not have,
such services.84
Potential barriers to competition in the supply of iTV services
Much of the current debate about APIs and their potential to function as barriers to
the supply of iTV services centres on whether STBs, containing APIs, are
“interoperable”. The interoperability concept appears to be being used in a number of
senses in this debate, including:
• the ability to be able to run an application or service on any appropriately
configured delivery platform;
• compatibility across interactive digital services and applications that require
consumer-side middleware to support them;
• the operation of receiving equipment between networks (supporting delivery
platforms); and
• the ability of networks to inter-work with other networks.
Broadly speaking, it appears that the concept is being used to embrace two distinct
‘families’of interpretability issues related to the development of iTV services, namely:
• content/ application - the ability of content/ application producers to author once
and to then port the content/ application between networks (or varying
capabilities) with the minimum amount of reauthoring possible and the reciprocal
ability of consumers to access content so authored;85 and
84
Quoted in “ITV Standards: The Interoperability Challenge”, Patrick Griffis, Nov- Dec
2002, at www.studio-systems.com/Broadcasting/NovDec2002/ITV%20Standards/94.htm.
85
In this respect, Recital 31 of the Framework Directive refers to interoperability of
digital iTV services at the consumer level, in order to ensure the free flow of
information, media pluralism and cultural diversity. It goes on to provide that it is
desirable for consumers to have the capability of receiving, regardless of the
transmission mode, all digital iTV services.
Ovum
November 2003
97
• network/ equipment – the ability of all receivers (STBs) to be successfully used on
all or most multiple networks and delivery platforms.
It is clear that authoring tools do not provide effective mechanisms to facilitate the
compatibility of all receivers with all platforms. As such, they will in no way facilitate
consumers switching between platforms (or operators) without being required to
acquire terminal equipment (i.e., STBs) that is compatible with the platform to which
they are switching.86 Equally, it is clear that authoring tools have the ability to
facilitate the cross-platform availability of iTV services, Information Society services
and other e-services, ensuring that they are available to as many European citizens,
over the broadest range of platforms and using the widest variety of end user
terminals, as is possible.
The adoption of a standardised API (whether mandated or on a voluntary basis)
would address both the content/ application and network/ equipment families of
interoperability issues identified above. However, the Commission may only mandate
a standard,87 under Article 17(3) of the Framework Directive, to “the extent strictly
necessary to ensure [interoperability of services] and to ensure freedom of choice for
users.” The balance of this chapter considers the potential role of authoring tools in
facilitating the production of interoperable iTV services, in an effort to provide
guidance as to the extent to which it is ‘strictly necessary’to mandate an API
standard to ensure service interoperability.
Issues associated with the porting of content/ applications
The fundamental issue in relation to the production of, and receipt of, iTV content and
applications is whether they can be made portable across platforms (i.e., can they be
authored in a manner that ensures that they are compatible with, and run on, at least
the vast majority of platforms (including their middleware). A number of factors tend
to reduce the extent to which content and applications can be ported, namely:
• transmission bandwidth (particularly the return channel);
• network integration;
• computational (or ‘processing’) resources in the STB;
• different APIs in the STBs;
• transaction processing; and
• linguistic and cultural issues.
86
Article 17 of the Framework Directive is directed at these issues.
87
It would appear that the existing variety of standards in the installed bases of
STBs across the Member States would render it difficult to permit standardisation by
individual Member States, without infringing the single market principles enshrined in
the Treaty.
Ovum
November 2003
98
This section describes each of these issues and reflects estimates of portion of
authoring costs that industry have indicated that they would attribute to each of these
issues.
Transmission Bandwidth
There are significant differences in (down-link) transmission capacity between
networks. In general, satellite tends to have the largest available capacity; terrestrial
has considerably less (due to both a smaller number of available channels and lower
transmission capacity per channel). In the UK, for example, Freeview carries 27 TV
channels, 4 interactive services and 12 radio channels. Sky carries 331 TV channels,
182 interactive services and 67 radio channels. As such, the balance between
providing services on TV channels (which can be locally stored on the STB for access
to a limited interactive experience) and interactive channels differs between platforms
(in addition to the clear raw capacity-related distinctions). Resizing content/
applications requires reworking of code, graphics and data to adjust to the amount of
available bandwidth, particularly when downsizing.
There are also not insignificant differences in the return paths. Cable STBs (with
cable modems) give a relatively high bandwidth, always-on connection. Satellite and
terrestrial STBs have used telecommunications modems, to date, providing a
relatively low bandwidth, dial-up connection. The characteristics of the different
return paths require content reauthoring to allow for the speed and availability
differences between platforms. Even more reauthoring is required to adapt services to
run without any return path (e.g., through storage of functionality that has some
interactive characteristics on the STB).
It is useful to consider the different approaches required for the different return paths
by reference to an example, i.e. a pay-for-play gaming application. Platforms without
a return path can provide such services only using ‘pre-paid’credits stored on the
STB (credited by the operator in response to a call from the consumer requesting
crediting) and paid for either in advance or billed on the next periodic bill. Platforms
with dial-up return paths could use the narrowband return path, with explicit
authorisation from the consumer to do so. Platforms with an always on return path
can interact with the consumer in real-time, requiring and obtaining consumer
authorisation at the outset. Adjusting content/ applications for this wide range of
alternatives requires significant adjustments to code, graphics and data.
The BBC has noted that cross-platform divergence is more a result of the different
technical capabilities of the different platforms than of the divergence in APIs.88
88
BBC R&D White Paper “Delivering Interactive DTV services to multiple target
platforms”, R. Bradbury, R. Cartwright and T. Steele, January 2003.
Ovum
November 2003
99
Network integration
The dynamic function of a number of iTV applications, including current news
headlines, weekly game shows with a ‘play at home’element, seasonal sports
coverage and services leading to the delivery of goods (including pizzas etc), uses
network infrastructure to access information necessary to trigger interactive elements.
Porting such applications requires integration work with the operator’s head-end
equipment and servers.
The costs of reauthoring content/ applications to address such integration issues
clearly vary on a case-by-case basis. However, Open TV has indicated to us that its
own experience with porting dynamic data indicates that infrastructure-related porting
costs alone frequently amount to 20% of the original application cost.
Computational/ processing resources
In the initial stages of DTV deployment, STBs with mid-range processing capacity
were deployed, at least in part to manage costs while volumes increased and
economies of scale in production developed. However, the range of STBs is
increasingly diversifying, from basic “zapper” STBs with little functionality to high-end
STBs (some incorporating sophisticated functionality like PVRs). The processing
capabilities of a selection of currently available STBs is set out in Figure 4.2, below.
Figure 4.2 Processing Capacity of STBs.
STB
Processor
RAM
ROM
Zapper
30 MHz+
1-2 MB
1-2 MB
MHEG-5
50 MHz+
4 MB
2 MB
OpenTV
50 MHz+
4-8 MB
4 MB
MediaHighway
50 MHz+
4-8 MB
4 MB
MHP Interactive Broadcast Profile
80-130 MHz+
8-16 MB
8 MB
MHP Internet Access Profile
150-200 MHz+
16-32 MB
16 MB
A white paper produced by Philips, Sony, Panasonic and Nokia89 states that, when
comparing like with like, the bill-of-materials90 cost difference between MHP
89
“The Costs of MHP in television receivers”, Philips, Sony, Panasonic, Nokia,
undated.
90
The bill-of-materials figures in the paper include, in addition to the costs of building
STBs with additional processing capability, the total costs of all components. It
should also be noted that the white paper (and the numbers that it contains) reflects
the technical characteristics (in terms of processor, RAM and ROM) set out in 3.4,
Ovum
November 2003
100
Interactive and Open TV/ MediaHighway Interactive was approximately €17 in 2001,
€4 in 2003 and would reverse (in MHP Interactive’s favour) to be €4 in 2004.91 It
goes on to indicate a bill-of-materials differential between MHP Interactive and a
basic MHEG-5 zapper box of approximately €55 in 2001, €37 in 2003 and €34 in
2004.92 The authors anticipate that memory and processor cost reductions will be the
main drivers for the cost reductions for all STBs. However, we note that the white
paper makes no provision for the costs of the intellectual property right licences
required. These costs are as yet unknown,93 as the relevant patent pool is not yet
operating.
The processing resources of an STB have a direct impact on the type and number of
applications that can run on the platform of which it is part. However, such resources
can not be ‘remotely’upgraded. Upgrade requires physical modification or
replacement of the STB. The likely costs of such switching out are described above.
APIs
In most cases, an API is effectively an operating system that incorporates both a
basic operating system and an application interface. In providing the interface
between the operating system and the applications running over it, an API enables,
limits and controls the iTV services that run over the platform of which it forms part.
Content and application producers must have access to the specifications of an API
and the authoring tools that are compatible with that API to be able to produce
content/ applications that will run on that API.
To the best of our knowledge, content/ application producers are not currently
experiencing difficulties in gaining access to the largely proprietary specifications of
APIs, to the extent that access is required to author content/ applications.94 Nor have
we seen evidence that the terms and conditions on which such access is provided
above, and include the cost of 56 kbps modems for the MHP and Open TV/
MediaHighway STBs.
91
The (unsubsidised) retail cost differences were anticipated to be a multiple of
three of these figures.
92
Again, the (unsubsidised) retail differences were anticipated to be a multiple of
three of these figures.
93
At this stage, the only agreed element is the $1 ceiling for each STB on the Java
components.
94
It appears that Article 18 of the Framework Directive provides a clear policy
support for the adoption and use of ‘open’APIs. However, the ‘open’concept is not
defined. While the APIs currently in use are proprietary (at least in part), it does not
appear that access to the rights necessary for authoring is restricted. In this context,
Article 18.2 requires API providers to make available the necessary authoring tools.
Ovum
November 2003
101
operate as barriers to competition. 95 Given that demand for iTV services is a function
of the content/ applications available, API rights holders (i.e., middleware producers)
do not have an incentive to restrict the availability of API specifications (at least those
elements necessary for authoring).96
There are currently at least seven APIs in use in Europe. The major deployed APIs,
as at March 2003, are set out in Figure 4.3. These APIs are not compatible, in that
STBs with one API cannot operate on another platform and services written for one
platform cannot run on others without a degree of reauthoring.
There is significant disagreement within the industry in relation to the portion of the
total cost of reauthoring content that can be attributed to the use of different APIs.
The breadth of the range of values attributed was highlighted at the October 2002
meeting of the DVB Commercial Module: Open TV, NDS and others estimated
attributable reauthoring costs to be in the range of 5 to 15% of total porting cost.
Philips, IRT and others estimated the costs to be almost 100%. Disagreement of this
magnitude must reflect, inter alia, consideration of fundamentally different types of
content/ applications (e.g., content originally authored to be portable, as opposed to
transaction-based content written specifically for a particular platform).97 In addition,
the higher estimates may also reflect the incorporation of costs related to the STB
(rather than to the API itself).
95
As such, it does not appear to us that there are issues relating to access, in most
Member States, that warrant the extension of the ‘fair, reasonable and nondiscriminatory’access standard to operators of APIs, under Article 5(1)(b) of the
Access Directive.
96
We note that Oxera reached a similar conclusion at page 27 of their February
2003 report “Study on Interoperability, Service Diversity and Business Models in
Digital Broadcasting Markets”.
97
In this respect, it is clear that the adoption of authoring tools or formats for new
content will have no impact on (and will not facilitate) the portability of existing content
authored without using such tools.
Ovum
November 2003
102
Figure 4.3 APIs currently in use
API
Platform
Liberate
UPC (Austria, The Netherlands, Scandinavia), ntl, Telewest
MediaHighway
Canal+ (Belgium, The Netherlands), Canal Satellite (France, Spain), Canal
Digital, Le Bouquet, Numericable, Tele+, Telenor Avidi
MHP
YLE, Helsinki Television
Open TV
Noos, TPS, UPC (France), FTC, Stream, AUNA, Casema/ Mediakabel,
Comhem, Senda, TeleDanmark, Via Digital, Sky, Cablecom
Betanova
Premiere
MHEG-5
Freeview
Microsoft TV
TV Cabo
Transaction processing
In relation to transaction-based applications (e.g., banking applications), platform
operators typically process transactions in relation to purchaser identification and
verification and delivery information. Any application using the platform’s resources
for such purposes must be adjusted to work with the platform’s database system and
the relevant transaction (and billing) model(s). In addition, given the sensitivity of
such applications in relation to a broad range of matters including data protection,
transaction security and fiscal regulation, extensive (and expensive) testing is
required.
The costs of porting transaction processing applications vary enormously. As such, it
is essentially meaningless to generalise in relation to such costs. However, we note
that a number of application providers have indicated that these costs can actually
exceed the total cost of developing the application for the original platform.
Linguistic and cultural issues
There are a number of linguistic and cultural issues that must be taken into account
when porting content/ applications. At a minimum, there are often translation
requirements. These may also require graphics, audio and text layout revisions, to
reflect the fact that messages in different languages require different amounts of
space (and time). For example, German requires more space than English.
Broadcasters and network operators also impose ‘look and feel’requirements that
serve not only to provide consistency and branding, but also to differentiate content/
applications from those offered by competing operators.
‘Reskinning’content/ applications to address these issues (i.e., changing its
appearance while preserving functionality) usually involves some reauthoring of code
and not insignificant work on textual and graphical characteristics.
Ovum
November 2003
103
In addition, there are cultural differences that render particular content/ applications
more or less relevant (e.g., the Commonwealth Games is relevant, at least to some,
in the United Kingdom, but is completely irrelevant in France).
Authoring Tools
The adoption of platform-independent content interchange formats appears to have
the potential to provide a mechanism for describing applications in a platform-neutral
manner. This is sometimes also characterised as authoring at a high level. Such a
high level approach was adopted in the development of HTML and Adobe’s PDF (i.e.,
formats suitable for describing the intended consumer experience via a set of
platform-independent elements including buttons, menus, formatted text, images,
tuners, key press and stream events and audio/ visual streams).
There are a number of current initiatives for the creation of standardised data
interchange formats supported by XML (an open standard that allows users to
precisely specify the content and structure of documents). XML is a text-based metalanguage that is human readable and can be easily transported using standard
protocols (including Internet protocols) over broadcast and return paths. It uses ‘tags’
to identify data elements (with the tags potentially defined for each document, as
appropriate). XML frequently uses Document Type Definitions (“DTDs”), carrying a
description of its format. Such DTDs can be used by independent groups of people to
interchange data, and to ensure that data to be interchanged is properly structured.
DTDs are a convenient way of packaging service descriptions as files or network
transactions, thereby providing a clear separation between content and design. As
such, services can be described in a purely abstract manner.
When XML is used in conjunction with cascading style sheets, presentation of content
can be controlled. The use of different style sheets allows a document to be
presented in a number of ways, without changing the document itself. A publication
engine manages the dependencies between content and templates. Content and
templates are bound together, to automatically generate platform-specific versions of
services, with platform-specific components. Design templates are expressed as
abstract components (e.g., position, size and colour scheme). Simple components
(e.g., text and bitmaps) map directly to presentable objects on STBs. Composite
components (e.g., menus) are collections of simple ones. The publication engine can
generate code representing objects available for each platform by mapping the
abstract service design to the range of objects available. As a result, the author of a
DTD description does not need to know how to code for a particular platform.
Instead, he or she relies on a library of pre-built components. Such systems are
illustrated in Figure 4.4 below.
Ovum
November 2003
104
Figure 4.4 Building iTV content/ applications
A number of XML application formats are currently in use, running on Open TV,
MHEG-5, DDE-1, Canal Plus, MHP 1.0 and HTML 4, including formats developed
both by middleware providers and platform operators themselves (e.g., Sky’s wtvml
solution). Most importantly, there are a number of initiatives to develop documented
international standards for a common application format which will work across the
widest possible array of standards, including:
• the SMPTE DDE-2 ad hoc working group – developing a standard for an
interoperable authoring format intended to bridge the gap between declarative
and procedural content by semantically defining the intent of the content author
using a language neutral methodology (which can be converted to particular
delivery standards).
• the so-called Portable Content Format sub-group (the “PCF Group”)98 of the MHP
Commercial Module – agreed on 16 June 2003 to work to define the commercial
requirements for a framework for creating iTV services for MHP and legacy
platforms with a minimum of reauthoring, acknowledging that there is likely to be
a remaining translation step to convert applications/ content from the authoring
format to the native capabilities of the device or network. It further acknowledges
that such a format should be, at least in part, transcodable into other formats, as
required.
Such authoring formats are unlikely to be appropriate for all genres of application.
However, we note that there does not appear to be a consensus among industry in
relation to identifying those genres for which standardised authoring formats are
unlikely to be appropriate. The PCF Group has noted that it needs to identify the
genres that would be best described in the new format, and to then consider scripting
and other extensions. It has indicated that it believes that the majority of popular
98
The Portable Content Format group was originally called the Common Content
Format group.
Ovum
November 2003
105
applications would benefit from the development of a portable format, but that arcade
style switch games, EPGs and other applications that are “written to the metal” of a
particular platform would not. However, we note that there is some disagreement
among industry players as to the types of content that are amenable to authoring in a
common format. For example, other sources have indicated that they believe that
EPGs would in fact be candidates for common format authoring.
During interviews held for this study, industry participants estimated that
approximately 80% of current iTV applications could be completely authored using
such a platform-independent language. Of the remaining 20%, sources consulted
during this study anticipate that some elements of the applications could be authored
using a common format.
Barriers to porting and authoring tools
Clearly, common authoring formats cannot address all of the issues reducing the
portability of content. In particular, they do not assist in any way in relation to those
‘barriers to porting’resulting from:
• transmission bandwidth;
• network integration (with the head-end equipment and servers on each platform);
• processing capacity (of STBs);
• tailoring transaction processing functionality to, and integrating with, each
platform; or
• the need to adjust services to reflect linguistic and cultural differences.
In addition, the adoption of common authoring formats will not reduce the cost of
reauthoring content/ applications that have already been developed on a bespoke
basis for one particular platform.
However, we note that:
• the component style of authoring of content/ applications that lies at the core of
mark-up languages and their platform-neutral output facilitates easier porting of
the integration-related issues set out above; and
• the network infrastructure and cultural and linguistic issues would also not be
addressable by the adoption of a standard API across all networks (whether
voluntarily or mandated).
However, it does appear that they have the potential to provide a means of
addressing the portability issues (of those that can be addressed) in relation to
approximately 80% of iTV content/ applications. As such, common authoring tools
appear to have the potential to facilitate both the development and provision of
Information Society services and other e-services, for as many European citizens,
over the broadest range of platforms and using the widest variety of end user
terminals as is possible.
Ovum
November 2003
106
The Study Team recommends that the Commission should :
• form a view, before July 2004, of the relevant market failures (if any) at particular
points in the broadcasting value chain, with particular focus on whether they
relate to reauthoring, standardisation or interoperability;
• explore fully the extent to which authoring tool-based obligations alone can
address the perceived barriers to competition. It would also be appropriate to
simultaneously consider the extent to which the published specifications under
Article 17 are ensuring the necessary degree of interoperability; and
• mandate standards or access (beyond the non-imperative language of Article
18(2)) only if these two measures prove insufficient. Before taking any such
steps, the Commission should consider whether other, less intrusive, remedies,
including more specific transparency (and publication) remedies, could address
perceived barriers to competition.
Ovum
November 2003
5
107
e-services
5.1 The current industry structure
There are six main components of the e-services supply industry:
• Consumer oriented ISPs who provide access, email, home page and content to
residential and small business users. Consumer oriented ISPs/portals include
Wanadoo, AOL, T-Online, BTopenworld, and Tiscali. ISPs in this category vary
considerably in the extent to which they deliver their own content to customers.
Some offer simple portals, search engines and starter packs; others, like AOL,
heavily invest in and promote their content.
• Backbone ISPs who provide IP transport and global Internet connectivity to
smaller ISPs. In Europe these are often vertically integrated with web hosting
service providers and buy global Internet connectivity from other backbone ISPs
like AT&T, Genuity, C&W, UUNet etc.
• Web hosting SPs. These companies host web sites for large and small
businesses. They run consumer facing applications such as e-shopping and ecustomer support for large corporates and serve closed user groups (CUGs) of
SMEs. Major companies here include Equant, BT Ignite, C&W, and T- Systems.
These companies have expertise in operating call centres and standardised, high
volume e-services.
• Managed application SPs who supply large corporates and CUGs of large
corporates and SMEs with back office applications such as HR, accounting,
supply chain management and CRM. These applications require more tailoring
and account management than those offered by the web hosting SPs. Large IT
services companies like IBM Global Services, CSC and EDS dominate this
sector.
• e-payments SPs. These SPs are orthogonal to the other SPs in the e-services
industry. They include the major credit card organisations and specialised multipayment services like Bibit.
• Managed technology SPs supply and manage the hardware, middleware and
software required by the managed applications SPs. These companies are often
the IT infrastructure outsourcing departments of the major IT services companies.
They also include software suppliers like Microsoft.
Figure 5.1 shows the value chain. We consider the first five segments of the eservices industry in this chapter and discuss the sixth in Chapter 6.
Ovum
November 2003
108
Figure 5.1 The value chain for the e-services industry
Consumers
Consumer
ISPs
Businesses
e-pay
SPs
Web
hosting
SPs
ASPs
Managed technology and backbone ISPs
5.2 Corporate e-services
Web hosting and managed application service providers supply the corporate market
with the bulk of their e-services. These service providers offer both online delivery
and management of business application services and infrastructure services. They
include services that have been developed to use the Internet to deliver a centrally
managed service that is accessed online by many customers (one-to-many). They
also include more traditional managed services that are being updated by IT service
companies to embrace the Internet as a service delivery channel (one-to-one).
The liberalisation of the telecommunications markets in Europe and the advent of the
Internet saw a surge in the provision of such e-services by both telecommunications
incumbents and new web-based service providers. From the beginning, most of these
new markets have been relatively competitive, with plenty of choice for end-users,
price-based competition and, relatively speaking, low barriers to entry. However, with
the collapse of investor confidence in telecoms, investment has reduced significantly
and the industry is currently going through a significant consolidation phase.
E-services in these segments vary from managed services like web-hosting, online
security services and online storage services to application services and business
process outsourcing (“BPO”). These services can either be provided in a
standardised, scaleable way to many end-users or customised to suit the process
requirements of each customer individually. The variety of service propositions is
illustrated by Figure 5.2. It is clear from this diagram that there is strong competition
here between the major telecoms operators – who are strongest in offering high
volume services- and the IT service companies – who are good at offering tailored
applications.
Ovum
November 2003
109
Figure 5.2 corporate e-service products and providers
Accenture
CSC
EDS
Unisys
SAP
Oracle
IBM Global Services
HP
T-Systems
BT Ignite
Managed
services
Microsoft
NTT Verio
Application
services
WorldCom
Service offering
Business
process
outsourcing
Content
services
Volume
solution
Premium
solution
Type of business model
Source: Ovum
This analysis suggests that there is strong competition in the provision of corporate eservices and no major barriers to competition requiring regulatory intervention. Those
interviewed for our study agree.
5.3 Backbone ISPs
Backbone ISPs provide other e-services players with global Internet connectivity.
This market segment has been particularly hard hit by the downturn in the market,
significant over-capacity and by resulting pressure on prices. As a result, many
backbone ISPs are in financial difficulties. A number of carriers in the related carriers’
carrier markets have gone out of business over the last couple of years. However,
this has not necessarily resulted in a reduction in capacity in the industry because
ownership reverts to the providers of the – often leased – assets. Figure 5.3 illustrates
the dramatic excess capacity for international bandwidth in Western Europe. The
graph gives an indication of available excess capacity in the market. The fully
upgraded capacity line largely represents fibre that has already been rolled out, but
has not been ‘lit’.
Ovum
November 2003
110
Figure 5.3 Pan Western Europe: supply versus demand
1,000,000
100,000
Gbit/s
10,000
1,000
100
10
1
2000
2001
2002
2003
2004
2005
2006
2007
Beginning of year
End user traffic
Bandwidth supply
Fully upgraded existing and planned supply
Source: Ovum
The emerging market structure sees the remaining telecommunications carriers
evolving into data/ IP carriers, capitalising on their strengths in providing simple,
standard online services (‘utility computing’) for the volume market (e.g., hosted
storage, security and messaging services) in addition to providing international
connectivity. Clearly, the excess capacity in this market limits the possibility of entry
and projected market revenues are not expected to generate a significant return on
the original investments.
Despite these difficulties, those responding to our study inquires believe that this
market is, and will remain, competitive and that regulatory intervention would be
harmful rather than helpful. For example ISPA, in the recent paper on the subject,
concluded that:
“Internet networks continue to develop in response to market forces. Unnecessary
regulation would involve significant costs, both directly and in terms of its effect on
investment in global, European and national Internet capacity and connectivity.“
One issue which has been the subject of considerable debate over recent years is the
US-centric nature of the Internet. Some governments and operators, especially from
the Asia Pacific region, argue that the terms on which their operators buy Internet
connectivity from US backbone ISPs are unfair and effectively transfer welfare from
the Asia Pacific region to the USA. The available evidence suggests that such
problems have all but disappeared, at least as far as the EU is concerned. The
development of local content and Internet exchanges within the EU, the use of
mirrored sites, and the global built out of networks into the EU by backbone ISPs
Ovum
November 2003
111
have all led to a much more equal balance of Internet traffic across the Atlantic over
recent years.
We conclude that:
• the market for Internet connectivity in the EU is and will remain competitive;99 and
• there is no need for regulatory intervention in this market.
5.4 Consumer oriented ISPs
Respondents have indicated that there are competition problems in the supply of
consumer oriented Internet services in that:
• the original business models of these service providers relied to a large extent on
revenues from advertising, e-business commissions and content supply. These
revenue streams have largely failed to materialise and the consumer oriented ISP
currently relies heavily on access revenues;
• initially these ISPs generated most of their access revenues by taking a share of
the dial-up Internet access revenues paid by the user to the fixed network
operator. However this source of revenue is now shrinking, as high usage
consumers switch from PSTN dialup access to DSL access;
• independent ISPs (i.e., those which are not subsidiaries of access network
operators such as fixed incumbents and CATV operators) are likely to disappear
unless:
-
they can quickly start to make money from content; and
-
they can make a reasonable profit on the resale of DSL services.
The prospects for independent consumer oriented ISPs making money from content
appear to be poor, currently. They depend on:
• whether consumers will pay for content accessed over the Internet. This is still
uncertain; and
• whether these ISPs have access to viable micro payment mechanisms. Credit
card payment is not especially effective for transactions of less than 10 Euros. In
addition, many Internet users are reluctant to use their credit cards. In addition,
the access network operators who provide the obvious micro payment
mechanisms are competitors of these independent ISPs.
The margin which independent consumer oriented ISPs make from reselling of DSL
services depends on the wholesale products that are available and their pricing. ISPs
99
The competitiveness of Internet connectivity has recently been acknowledged by
the Commission in its Recommendation on Relevant Product and Service Markets
within the electronic communications sector susceptible to ex ante regulation in
accordance with Directive 2002/21/EC of the European Parliament and of the Council
on a common regulatory framework for electronic communication networks and
services.
Ovum
November 2003
112
claim that wholesale DSL priced on a retail minus basis does not provide an
adequate margin but that bitstream DSL priced on a cost plus basis does offer such
margin.
NRAs face a difficult choice here. If they try to preserve competition from independent
consumer oriented ISPs by setting cost based prices, they could significantly reduce
further investment in broadband infrastructure. In selecting the appropriate regulatory
response, NRAs will need to weigh:
• the economic value that competition between such ISPs and those affiliated to
the access network operators brings against
• the reduction in basic infrastructure investment and competition which a costbased pricing rule on DSL bitstream access might create.
It is difficult to estimate technology and demand uncertainties correctly. These risks
however have to be taken into account when setting cost based prices for products
like DSL bitstream. These issues are discussed in Section 2.6 (Constraint 4).
The policy decision relating to the regulatory support to be provided to independent
consumer oriented ISPs could be an important one in determining the future level of
infrastructure competition within Member States. So we recommend that the
European Commission studies it in more detail with the aim of developing guidance
for NRAs – perhaps in conjunction with the ERG.
Our investigation also highlights two other barriers to competition which consumer
oriented ISPs face:
• many small service providers are deterred from web hosting from fear of being
sued by Hollywood studios and other IPR holders. The EU has acknowledged the
need for action through a “notice and take down”mechanism whereby a service
provider who takes down offending content when notified is protected from
prosecution. However, there is still a need to prescribe the process in detail
(whether or not in the manner that it has been specified in the USA); and
• there are problems of discrimination by incumbents, between the manner in which
they treat their own consumer oriented ISPs and those of rivals. Given the large
market share of the consumer ISPs owned by the incumbents, as set out in
Figure 5.4, there is clearly a need for regulatory diligence. To maximise
competition, it is important that NRAs ensure that incumbents, as the main
access network operators, interact with their own ISPs and independent ISPs in a
non-discrimination fashion. It would appear that this may not have always
occurred:
-
in the supply of dial-up access in some countries the ISP affiliated to the fixed
incumbent launched retail products where the charges are set on a flat
monthly basis without its rivals being able to buy flat rate Internet access call
origination (FRIACO) products on equivalent wholesale terms
-
ISPs have also complained that incumbents discriminate in favour of their
own ISPs in relation to the provision of Internet connectivity. According to
one CLEC, incumbent operators discriminate between third party ISPs and
Ovum
November 2003
113
their own downstream ISPs. They configure their own service offerings so
that Internet traffic is handed over to the incumbent’s IP backbone at a point
as close to the originating customer as possible. However, they provide
unaffiliated ISPs only with a higher cost access solution which involves use of
circuit switched networks and ATM backhaul. As a result, this CLEC claims
that its costs are significantly higher than those of the incumbent’s affiliated
ISP.
We recommend that the European Commission should:
• review the need to issue detailed guidance on “notice and takedown”procedures
to ISPs; and
• continue to investigate the complaints of discrimination in access provision which
independent ISPs bring against incumbents.
Figure 5.4 The market share of the incumbent’s consumer ISP subsidiary
Incumbent
Subsidiary
Market share
Deutsche Telekom
T- Online`
55%
France Telecom
Wanadoo
43%
Telefonica
Terra
41%
BT
Openworld
22%
Source: Ovum
5.5 e-payment issues
Most e-commerce payments are currently made through financial services institutions
regulated under banking legislation. E-commerce uses card-based methods of
payment (e.g., credit cards, debit cards, charge cards). These methods of payment
100
are suitable for reasonably large value payments, over €10, for example. The
market for the supply of such payment services is reasonably competitive and there
are no major issues requiring regulatory intervention on competition grounds.
However, there are practical problems:
• in some countries, consumers are reluctant to use credit cards or debits cards for
making payments online; and
• for low value transactions, such as micro payments for web content, credit cards
are not cost effective.
100
This value changes depending on typical usage of payment cards. US
consumers regularly use credit cards for transactions of a few dollars, whereas
German usage of cards is much less common. Other cultural factors are also
relevant, such as attitudes to security, trust and cash alternatives.
Ovum
November 2003
114
5.5.1 Micropayment systems
We have considered, in more depth, the regulatory problems that have been encountered
in the development and delivery of effective micropayments systems. The remainder of
this chapter:
• briefly describes how the different micropayments systems work, including the
transaction types and supply chain;
• identifies the key existing regulatory instruments that potentially apply to
micropayments systems;
• identifies the extent to and manner in which such regulatory instruments regulate
elements of particular forms of micropayments systems;
• discusses the manner and extent to which the application of such instruments appears
to affect the operation of micropayment mechanisms; and
• raises some proposals relating to the scope and application of such regulatory
instruments to micropayments.
What are micropayments?
A micropayment is a low-value electronic purchase. There is some variation in
relation to the quantum of micropayments. However, market evidence suggests that,
in most Member States, the value at which transaction costs render credit card and
other macropayments systems inappropriate is approximately €10. This level is
higher than that found in the United States, where credit and debit cards are routinely
used for transactions of $5 value or less. The Study Team has adopted the range
shown in Figure 5.5, reflecting the views expressed by industry and analysts.
Figure 5.5 A definition of micropayments
Macro
Micro
€0
€1
€10
€100
€000
Source: WirelessInternet@Ovum
Micropayments are made for ad hoc digital and physical content. Although the concept of a
micropayment was conceived in the fixed Internet world, micropayments are quickly
becoming important in the evolving wireless Internet services market. This is illustrated by
the various types of content that can be purchased wirelessly with micropayments.
Figure 5.6 lists key content types.
Ovum
November 2003
115
Figure 5.6 Key content types
Digital content
Ringtones
Icons
Gaming services
MP3 files
Premium information (such as real time stock quotes)
Location-specific information (such as a taxi locator)
Physical content
Ticketing services (for the movies, the theatre and so on)
Vending machines (‘dial-a-coke’type services)
Retailer co-operative services (Starbucks, McDonalds and so on)
Methods of payment
A variety of payment methods exist to enable micropayment transactions, including:
• metered payments
• optimised card payments
• e-cash
• pre-paid value.
These alternative methods are generally used in the contexts shown in Figure 5.7.
Figure 5.7 The domain of different payment methods
Payment method
Level of current
deployment
Best suited for
physical or digital
content?
Best suited for micro
or macro payments?
Metered
Medium
Both
Both
Optimised card
payments
High
Physical
Macro
E-cash
Low
Digital
Micro
Pre paid value
Medium high
Digital
Micro
Ovum
November 2003
116
Metered payments
A metered payment is one which is charged directly to a customer bill, such as a telephone
bill. Metered payments exploit existing payment infrastructures to bill the consumer for
purchased goods and services. Metered payments can be supported by anyone who has a
billing relationship in place with the consumer.
Optimised card payments
In the fixed-Internet world, the most obvious method of e-payment has been the credit card.
However, debit and credit cards do not provide an economic payment mechanism for
micropayments, given the cost (both relative and absolute) to merchants represented by
the charges imposed on the merchants for each transaction. Today, there are several
enhancements available to traditional credit card payments that are also significant.
• Virtual cards. A virtual card is a piece of thin-client software that is embedded in
a desktop, mobile phone or other electronic device. A personal identification
number (PIN) is used to link the consumers identity to the virtual card. The
buyer’s card details are held securely by the issuing bank for the virtual card,
rather than on the buyer’s device. The issuing bank verifies the PIN and thus the
identity of the buyer, and subsequently authorises payment to the merchant.
• Wallets. A wallet is an application that stores payment information (such as credit
card details), applications, virtual cards, data or links to remote functionality in a
desktop PC, a mobile phone, a smart card or an online server. Such information
typically includes the name and billing address of the buyer, plus details of one or
more methods of payment. Wallets were inspired by the time-consuming entry of
identical billing information into multiple websites. Wallets may be stored on the
device itself, or they may be network-based and securely accessed from the
device.101
E-cash
E-cash, also known as an electronic purse, is an attempt to create ubiquitous, cash-like
capabilities in an online payment method. It was conceived for high-volume, low-value
transactions such as micropayments. The concept of e-cash is similar to prepaid solutions,
but differs in that:
• e-cash is widely exchangeable, whereas prepaid solutions are reserved for a restricted
range of purchases
• e-cash can always be converted back into physical cash or be deposited back into a
bank account. Prepaid anticipates payment for something, and it is difficult to get your
money back once you have made the initial investment.
E-cash has failed to gain wide acceptance, either in the physical world (typically based on
smart cards) or in the fixed-Internet community.
101
See, for example, http://www.mobiletransaction.org/documents.html for further
details.
Ovum
November 2003
117
Pre-paid value
Pre-paid value is stored on prepaid cards issued by mobile operators. Purchases made
against such value exploits the existing pre-payment mechanism to bill consumers for
purchased goods and services. Given the relatively low average value on prepaid cards,102
pre-paid value is best suited for micropayments.
The different players in the value chain
The emergence of micropayments has created new roles. In the view of the Study Team, a
division between service providers and solution vendors (associated with the various
mechanisms identified above) is beginning to emerge, as shown in Figure 5.8 below.
Figure 5.8
The value chain for micropayment services
Technology enabler
Payment
solution
vendor
Payment
Service
Provider
Payer
Payee
Source: Ovum
Technology enablers
Companies such as Encorus and iPin are specialist providers for electronic payment
solutions. Given how little has been launched, it is too early to see clear leaders in the field.
Some service providers are moving towards this area including, for instance, Paybox.
Payment solution vendors
Vendors are already on the market trying to sell different solutions for electronic payment.
As yet, there is no single solution which has established itself as the market leader.
However, much effort is going towards the creation of a common approach.
102
The Study Team understands that the average pre-paid value currently
maintained by customers is between € 10 and 15.
Ovum
November 2003
118
Payment service providers (PSPs)
The area that is drawing most of the commercial speculation is the payment service
provider (PSP) area. There is still uncertainty as to who is set to become the primary
PSP(s). The PSP adds value with services that include:
• security
• transaction aggregation
• currency exchange
• billing
• account consolidation.
Hosted service providers that have the competence and experience to become PSPs are
developing products for the wireless Internet. Mobile service providers with skills in
security, billing and transaction aggregation services are well positioned.
5.5.2 Key players in the micropayment value chain
As Figure 5.9 below illustrates, mobile phone companies are becoming the leading mobile
PSPs (e.g., i-Mode and Vodafone). They appear to be better positioned in Europe than
financial players such as Visa. Market evidence suggests that many banks and other
financial services organisations have effectively abandoned the provision of micropayment
transaction services. There have been a number of relatively high profile ‘failures’in the
provision of such services by financial institutions (e.g., Mondex and Visa Cash). A small
number of systems continue to operate successfully, primarily for larger value transactions
(e.g., Proton). It may be that the transaction costs (including those imposed on third parties
accepting stored value as consideration) of such systems could be lowered, so that they
could move more effectively into the micropayment space.
Other specialist players, such as Paybox, are starting to concentrate on being solutions
vendors and are moving away from service provision. Paypal, now part of the Ebay group,
is a leader in terms of breadth of service functionality and service provision, though mostly
in relation to the fixed Internet.
Although ASPs or ISPs could decide to become payment solution vendors, it appears that
there is more opportunity for them to leverage their expertise in application development
and integration to support back-office processes. Creating seamless links between
customer-facing applications and back-office settlement systems is difficult, but an ASP/ISP
with experience and expertise in developing and supporting this functionality will be well
positioned to do so.
Ovum
November 2003
119
Figure 5.9 Key players in the micropayments services sector
Followers
Leaders
High
PayPal
Vodafone
Level of service provision
Visa
Service
focus
i-Mode
MobiPay
paybox
IBM
Enablers
Encorus
iPin
Low
Low
High
Breadth of functionality
Source: Ovum
Micropayment market development
Micropayment mechanisms in the fixed Internet environment have not been a success to
date. However, they are much better positioned to succeed in the wireless space, not least
because consumers are accustomed to paying for wireless data services (the best example
being the success of SMS). Most observers believe that this is because consumers are
more appreciative of the contextual manner in which they require the use of ‘mobile’
services. Specifically, when a consumer is ‘mobile’, they are more willing to pay for the
value inherent in just-in-time delivery.
The wireless Internet is a new playing field, and most of its current customers are early
adopters. Consumers (as well as business users) are not used to receiving free services
over their mobile handsets, and they are particularly receptive to paying for premium
content delivered on a just-in-time basis. In terms of acceptance and adoption, industry
data suggests that wireless micropayments have made significant progress in the last year.
The success of wireless micropayments may also have an impact on payment methods
over the fixed Internet. Wireless micropayments may provide a ‘back door’for the
emergence of a ubiquitous micropayment scheme for the fixed Internet. (e.g., Vodafone’s
m-Pay Bill permits the billing of purchases over the fixed Internet to the customer’s mobile
bill).
It does not appear that it will be difficult to create an incentive for third-party content
providers to join a micropayment mechanism. There are two main incentives:
Ovum
November 2003
120
• most content providers have grown from the print media industry. Many have had
success in moving their business to an online format. Micropayment mechanisms
represent a means to derive real revenues from such offerings.
• the immediate impact to the bottom line. Some PSPs with a pay-per-use model pay
their third-party content partners on a monthly basis. The resultant revenues, improving
bottom line performance, can have a rapid and significant impact.
It is too early to say how the market will evolve. Most ECS providers are closely examining
the regulatory implications of offering payment services. Banks are also looking for signs of
increased competition for a share of the ‘customer wallet’.
5.5.3 Micro-payments - the key regulatory instruments
There are a number of key regulatory instruments that potentially apply to various providers
of micropayment systems and aspects of particular micropayments systems, including:
• the E-money Directive;103
• the Electronic Signatures Directive;104
• the Money Laundering Directive.105
We summarise, in Annex A, the key provisions of each of these instruments. Other
instruments, including those relating to tax and consumer protection, may impact on the
design and operation of micropayments but do not do so in any manner that is particular to
micropayments (when compared to any other payment type). As such, these issues are
not addressed in the analysis of this chapter.
5.5.4 Micro-payments - the scope of the E-money regime
The key current regulatory issues arising from the legal framework governing the provision
of micropayment systems are centred on the scope and appropriate interpretation of the
“electronic money”concept. The E-money Directive only applies to stored value that is
stored on an electronic device (including cards containing chips) and is accepted by
103
Directive 2000/46/EC of the European Parliament and of the Council of 18
September 2000 on the taking up, pursuit of and prudential supervision of the
business of electronic money institutions. In addition, Directive 2000/12/EC of the
European Parliament and of the Council of 20 March 2000 relating to the taking up
and pursuit of the business of credit institutions, as amended by Directive
2000/28/EC, might also apply to micropayment system providers that are e-money
institutions.
104
Directive 1999/93/EC of the European Parliament and of the Council of 13
December 1999 on a Community framework for electronic signatures.
105
Council Directive 91/308/EEC of 10 June 1991 on prevention of the use of the
financial system for the purpose of money laundering, as amended by Directive
2001/97/EC of the European Parliament and of the Council of 4 December 2001.
Ovum
November 2003
121
undertakings other than the issuer of the value. Only pre-payment mechanisms are
covered by the E-money Directive. As such, the following types of micropayment
mechanisms will not fall within the scope of the E-money Directive, namely:
• systems that do not involve storage on an electronic device of monetary value (e.g.,
post-paid systems which do not entail the storage of value and other “metered
payment”systems, as described in section 5.5.1, above);
• systems in which only the issuer of the monetary value accepts the value as a means
of payment.
The practical effect of the application of e-money regulation to mobile payment systems
might well lead to the differential treatment of pre-paid and post-paid customers (to the
detriment of the services made available to consumers who are pre-paid customers)
despite the fact that both mobile subscribers and operators might have an interest in
ensuring that the same set of services can be made available to both categories of
customers.
Does pre-payment to ECS providers constitute “electronic money”?
The real issues, for ECS providers at least, are the extent to which and context in which the
E-money Directive applies to prepaid systems. More particularly, these issues stem from
the regulatory implications of the characterisation of pre-paid systems as “electronic
money”. It is essential that the inherent differences between such payment systems that
are only accepted by the issuers of value and those that are accepted by a range of third
parties are acknowledged and reflected in their regulatory treatment. The value “trade off”
between a ubiquitous micropayments system with relatively high transaction costs
(including regulatory costs) such as Proton and the smaller user base of a ‘closed’system
with potentially lower costs (such as pre-paid cards for mobile calls) is a commercial matter
for the entities establishing such systems. Regulatory uncertainty and excessive costs
imposed by legislation should not become factors to be considered in such decisions.
There are a number of issues that have arisen from various elements in the definition of
electronic money that are the source of uncertainty and inconsistency (in either
implementation or application) in this respect:
• what constitutes “acceptance” by an undertaking other than its issuer. More particularly:
-
is it appropriate to consider arrangements between the ultimate suppliers of
products and services and the entity entering into transactions with consumers
when identifying the entity that accepts stored value;
-
should the nature of a service or product acquired be relevant in determining
whether stored value is accepted by an entity other than the issuer;
-
should the nature of the deliver mechanism for products or services acquired be
relevant in determining whether stored value is accepted by an entity other than the
issuer; and
• the points at which, and methodology by which the distinction between electronic
money and credit transfers within a closed accounting system should be drawn.
Ovum
November 2003
122
When is value accepted by an undertaking other than the issuer?
A crucial issue for ECS providers currently operating or seeking to operate pre-paid
micropayment systems is the circumstances in which the system will be found to create
monetary value that is only accepted by the issuer, and not by third parties. Unfortunately,
the E-money Directive itself gives little assistance in relation to the legislative intent behind
the scope of the concept “accepted as means of payment by undertakings other than the
issuer”. As such, it is not clear how arrangements under which the issuer of the value is
the only entity that accepts the value, even though a third party provides the product or
service sought, are to be assessed. There are essentially three models under which
customers can acquire access to third-party services using a fixed or mobile ECS:
• the ECS provider buys (i.e., acquires “title”itself) the good or service from the thirdparty and resells it to its customers;
• the ECS provider has an arrangement of some form with the third party provider under
which the ECS provider accepts payment and, accordingly, the risk on behalf of the
third party provider; or
• the ECS provider provides connectivity to the consumer to enable them to request
supply from the third party itself, the third party takes the risk of non-payment and bad
debt, and accepts payment itself.
It appears, contractually, that, in the first instance, the prepaid value is accepted by only the
issuer. As such, it does not constitute electronic money. In contrast, the third approach
would appear to constitute electronic money, where a third party service or product provider
accepts the stored value as payment.
The relationships between service providers and suppliers
A key issue in the regulation of micro payments services is the relationships between
the entity entering into transactions with consumers (the service providers) and the
ultimate producers and suppliers of products and services.
The uncertainty as to the scope of Article 1(3)(b)(iii) is centred on arrangements under
which the ECS provider has an arrangement with the third party provider under which the
ECS provider accepts payment (and risk) on behalf of the third party provider (i.e., the
second scenario described above). There are a number of interpretations open to
consideration. More particularly, the following alternative approaches have been adopted
or considered to date across the Member States:
• Approach 1: that payments made to the ECS provider in relation to which it bears the
risk are payments (using the monetary value stored on the device) to the issuer of the
value. As such, they are not electronic money;
• Approach 2: that payments made to the ECS provider for digital services that are
supplied using the same ECS as that used to request the services are payments for the
supply of a single ECS supplied by the ECS provider (i.e., the issuer of the stored
monetary value). As such, they are not electronic money;
• Approach 3: that payments made to the ECS provider for either digital services that
are supplied using a different ECS than that used to request the services or for
Ovum
November 2003
123
products are payments for a separate and distinct service. The ‘flow of value’between
the ECS provider, the customer and the third party supplier is considered to constitute
the issue of electronic money.
The first approach, on one hand, and the second and third approaches, on the other, are
based on diametrically opposed interpretations of the meaning of “accepting payment”.
Approach 1
The first approach looks only at the identity of the entity receiving payment from the stored
value device. It takes no account of the nature or form of the arrangements that may be in
place between that entity and the third party content or product producer, of the service or
product supplied or of the ultimate means of supply. In essence, it focuses only on the
direct payment relationship between the service (or product) acquirer and the entity
receiving payment from him or her.
Arguably, this approach is entirely appropriate, since the identity of the ultimate supplier
could be considered to be irrelevant in relation to electronic money regulation that is
expressly intended to facilitate the creation and use of an electronic surrogate for coins and
banknotes (i.e., an alternative to legal tender). Once the value has been accepted in
payment, subsequent revenue flows between entities in the supply chain, and the nature of
the contractual relationships governing related arrangements between the entity accepting
payment from the purchaser and the third party (e.g., the nature of the arrangements
between the ECS provider and the third party supplier) are, arguably, irrelevant. The
appropriateness of making a clean regulatory break between the acceptance of the
electronic money and value transfers elsewhere in the supply chain is supported by three
elements:
• there is no direct payment obligation between the customer and the ultimate producer
or supplier of the product or service;
• the amount paid by the customer for a product or service is different to, and is not
necessarily directly related to, the amount received by the ultimate producer or
supplier; and
• the ultimate producer or supplier is usually unaware of whether the sum that it receives
in payment from the entity below it in the distribution chain relates to a service or
product that was paid for using stored value or was post-paid.
However, this approach could facilitate the circumvention of the regulation of electronic
money by allowing entities issuing stored value (including ECS providers) to enter into
arrangements with any third party provider under which anything could be acquired. The
Study Team notes that the potential scale and breadth of such arrangements, and their
resulting potential impact on prudential policy, are constrained where the issuer of the value
must assess and assume the risk inherent in stored value transactions relating to the
particular products or services. It appears that a workable solution might be to make it
clear that the “acceptance” concept refers to the principal in the transaction with the
consumer. Where the ECS provider is merely a billing agent for the ultimate principal in
such transactions, rather than acting as the principal in relation to the consumer-facing
transaction, it would not be “accepting”the value.
Ovum
November 2003
124
Approach 2
The second and third approaches essentially presuppose that it is necessary and
appropriate to consider both the nature of the service for which payment is made and its
delivery mechanism in identifying the entity that has ‘accepted’the stored value. It is not
entirely clear to the Study Team why either prudential or consumer protection policy
requires the ‘reading in’of these elements to the definition of electronic money.
More particularly, the second approach identified introduces the following three additional
elements into the circumstances in which electronically stored value will not be regulated as
electronic money:
• a temporal requirement that the service(s) acquired be acquired virtually
simultaneously with the originating request for provision from the customer;
• a requirement that the value is only used to acquire digital services that are capable of
delivery over the same form of ECS as the originating request for provision (i.e.,
excluding other services and all forms of products); and
• a requirement that the service(s) acquired are delivered over the same ECS as the
originating request.
These additional elements are clearly not part of the expressly defined electronic money
concept. The approach appears to hinge on the view that, in relation to digital ECSs (and
the content that might be inherent to such services) provided in the prescribed manner, the
stored value does not constitute a payment instrument (i.e., an alternative to legal tender).
It appears to amount to a pragmatic attempt to differentiate between value paid in respect
of ECSs provided by ECS providers (i.e., their core business), on the one hand, and value
used for other products and services in relation to which there is no connection with the
business of the ECS provider (other than as a payment mechanism), on the other.
This approach arguably reduces the potential for circumvention of electronic money
regulation through the widespread entry into arrangements with third party providers of
products and non-ECS services by ECS providers (discussed above). As such, it could be
seen as an attempt to identify the circumstances in which an ECS provider is most likely to
be acting as a principal in transactions with consumers. However, it relies on an underlying
assumption that is technology-determinant, in that it would clearly preclude the provision of
content or sophisticated ECSs across multiple delivery platforms where the ultimate
delivery platform differs from that used to initiate the supply transaction.
Approach 3
The third approach described above represents an extension of the second, in that it ‘looks
behind’the relationships that might exist between the ECS provider and third party service
or product providers. However, in doing so it is virtually writing out of the electronic money
definition the relevance of the issuance and acceptance of the stored value. It focuses on
the transfer of value, whatever form it might take, between a range of entities. It does not
acknowledge that the quantum of the value transferred at different points in the distribution
chain will almost certainly differ or that the billing mechanism adopted (i.e., pre-paid or
post-paid) may be opaque to the third party provider.
Ovum
November 2003
125
Credit transfers vs value circulating as bearer instruments
The Commission itself has raised an alternative view of the appropriate regulatory
characterisation of arrangements under which the issuer of stored value accepts payment
(and risk) on behalf of a third party provider, in the course its current review of the existing
regulatory structure for retail payment instruments. This review appears to be focused on
instruments that are provided and used as alternatives to legal tender. Of most relevance,
in the present context, the working document distinguishes between specific purpose
payment instruments which, inherently, have limited usability (e.g., can only be used to pay
for goods purchased from a particular retailer) and those that are true general-purpose
payment instruments.106 The working document sets out a view in relation to “pre-paid and
post-paid small value accounts used for third party payment services” that can be
summarised as follows: the E-money Directive applies to pre-paid systems where the
monetary value issued circulates on a bearer instrument. Systems that operate as
micropayment systems that are in reality credit transfers in a centralised account system,
rather than real bearer instruments, do not constitute electronic money.
Conclusions
We believe that the scope of Article 1(3)(b)(iii) should be clarified. In particular, it would be
helpful to:
• define the ‘accepted by undertakings other than the issuer’concept, to ensure that it is
not interpreted in a manner than brings specific purpose instruments and credit
transfers within closed systems within its scope;
• distinguish between acceptance of payments by entities as agent or principal, ensuring
that the development of innovative payment mechanisms that do not constitute general
purpose payment instruments is not hampered;
• review existing approaches to the implementation of Article 1(3)(b)(iii) to ensure that
they are not technologically prescriptive and do not, in effect, favour ECS platform
operators over other entities that may operate multiple-platform payment mechanisms;
and
• review existing measures against the characteristics that are essential for a functional
micropayment mechanism (e.g., that it be automated, secure and simple for the
customer), to ensure that regulation is not rendering micropayment mechanisms
unworkable.
The working document suggests that the Commission may not be minded to provide a
107
“literal interpretation” of the electronic money concept.
An alternative would be to
provide guiding principles. It would appear to be appropriate for the Commission’s ongoing
106
The Meeting Document concerning a New legal Framework for Payments in the
Internal Market. MARKT/4013/2003.
107
Annex 1 to the Meeting Document concerning a New legal Framework for
Payments in the Internal Market. MARKT/4013/2003.
Ovum
November 2003
126
deliberations concerning the appropriate course to take into account the extent to which
existing practice in many Member States currently does not reflect the ”bearer instrument”
interpretation (summarised above) of the status of existing pre-paid micropayment systems.
5.5.5 Implications of applying the E-money regime
There are a number of regulatory obligations imposed under the applicable instruments that
apply to entities found to be issuers of electronic money, whether as EMIs or as credit
institutions, that impose not insignificant burdens on such entities, including:
• the requirement to structurally separate EMI activities from the remainder of such
entities’other business activities;
• the requirement to keep funds related to EMI activities (e.g., stored value) within the
structurally separate entity conducting EMI activities;
• the imposition of the prescribed minimum funding for the EMI entity;
• the imposition of the investment restrictions in the E-money Directive on the EMI entity
(see Annex A);
• the requirement that pre-paid sums be redeemable;
• the reduction in the likelihood that the preconditions for a waiver under the E-money
Directive will be met (see Annex A);
• the compliance with supervisory costs of the EMI and credit institution regimes; and
• the taking of steps to adequately identify customers engaging in non-face-to-face
transactions, to meet the requirements of the Money Laundering Directive.
To the extent that the “electronic money”concept is implemented in an over-inclusive
manner, issuers of funds that are held in a form that does not represent a bearer instrument
are subject to such obligations.
Separation of electronic money
There are a number of quite significant practical difficulties that arise from the requirement
that liability referable to issued value be stored separately, particularly in light of the
proportion of the mobile ECS customers across the Member States that are pre-paid
customers.108 In jurisdictions in which the use of value on pre-paid cards for services other
than basic voice services is considered to be electronic money, mobile ECS providers are
unable to use stored value made through prepayment in the ordinary course of their
business. In any instances this removes a large portion of potential cash flow from the
ordinary course of an operator’s business.
It should be recalled that the requirement to keep electronic money stored value separate
to all other funds can effectively require the separation of all pre-paid value, including that
108
There are no pre-paid billing systems in Finland. However, the proportion of prepaid mobile customers in the other Member States ranges between 60 and 90% of
the total customer base.
Ovum
November 2003
127
which is not electronic money, because of the potential for stored value to be used for any
purpose, at the choice of the customer. For example, a customer can elect to use any
proportion (or all) of stored value for voice services, for PRS, for digital content delivered
over the same ECS connection over which it was requested, for digital content delivered
over an ECS connection other than that over which it was requested or for physical goods.
At the time that the customer buys the prepaid card on which the value is stored it does not
identify the purpose for which the stored value will be used. We understand from industry
that consumer surveys indicate that there would be very strong resistance from customers
to any attempt to require them to specify in advance the intended use of stored value, to
acquire different pre-paid cards for different purposes or to acquire cards which themselves
limit the services (and products) with respect to which particular value can be used.109 In
addition, any attempt to introduce any such distinctions would undermine the utility of prepaid mechanisms as simple, flexible and consumer-friendly billing systems.
These issues raise a real practical problem for ECS providers. One approach would be to
treat all stored value as electronic money in the first place, only reclassifying funds that are
not electronic money when they are used and it becomes clear that they are not electronic
money. Alternatively, such operators could treat all funds as ordinary non-electronic money
until they are used and it becomes clear that they are electronic money. In either case, the
‘trigger’that allows the appropriate classification of funds is the use of the funds. However,
electronic money regulation is expressed to apply as and when the value is stored. Any
pragmatic departure from classifying pre-paid value as it is stored amounts to a departure
from, and a potential breach of, electronic money regulation. However, any attempt to
impose electronic money regulation on the basis of assumptions as to the future use of
stored value is essentially artificial.
We believe that the resolution of many of these problems will be assisted by the
clarification of, and the provision of Member State implementation guidance in relation to,
the appropriate scope of “electronic money”.
However, while such clarification is being achieved, consideration should be given to, and
guidance provided in relation to, the appropriate and proportionate manner in which liability
derived from pre-paid value issued as part of a micropayment mechanism should be
separated from other issuer funds.
Separation of EMI activities
The characterisation of value stored on pre-paid billing mechanisms operated by ECS
providers as electronic money requires that the billing activities related to EMI activities be
structurally separated from other activities that are not closely related financial and nonfinancial (including operational and other functions ancillary to issuance) services. This
clearly requires separation from the core business activities of ECS providers (e.g., the
109
In this context, we note that any technological solution permitting consumers to
alter the “split”of value would require a means to communicate such alteration to the
value issuer, to allow the value issuer to move the related liability into or out of its emoney accounts.
Ovum
November 2003
128
provision of ECSs to customers). As such, it undermines a number of the key efficiency
drivers that would otherwise be behind the provision of billing mechanisms by ECS
providers.
The minimum funding level
The minimum funding requirement effectively requires ongoing access to potentially quite
significant funds in circumstances where the quantum of such funds is calculated by
reference to purchases of goods or services which the ECS provider has either supplied
itself or has elected to assume the risk that supply would be effected by entities of its own
choosing. As such, the prudential risks that justify the imposition of such a requirement - in
relation to stored value that is truly a bearer instrument to protect against the risk that value
will not be ‘honoured’by the issuer when it the customer seeks to redeem it - are not
present. It would appear that the obligation is unnecessary is such circumstances,
particularly where only the issuer itself (e.g., the ECS provider) is the entity accepting the
value stored in the first instance.
Investment restrictions
The investment restrictions are imposed on EMIs to ensure that they are sufficiently and
appropriately financially stable. It is not clear that these are necessary or appropriate in the
context of pre-paid micropayment mechanisms operated by ECS providers, given the
ancillary nature of such micropayment mechanisms to the core business of ECS providers
(with the effect that any entity established to hold liability for stored value will not be used
as an investment vehicle for anything other than the liability held) and the current
uncertainty in relation to identifying when funds actually become electronic money. The
prudential policy grounds behind such a requirement are clear. However, it is not clear that
ECS providers, given their prudential profiles, warrant this additional level of supervision. It
is also not clear that there are real consumer protection grounds for imposing this
requirement, in light of the comfort that the financial position of ECS providers should
provide to consumers in relation to the security of the small sums that they store in pre-paid
mechanisms.
We believe that many of the problems associated with separation of EMI, minimum funding
levels and investment restrictions will be resolved by the clarification of, and the provision
of guidance in relation to, the appropriate scope of “electronic money”, as generally
accepted bearer funds.
Redeemability
Article 3 of the E-money Directive requires electronic money to be redeemable, at the
request of the bearer. Many of the ECS-related prepayment systems are not redeemable
or were not designed to be redeemable. In view of the average quantum that customers
Ovum
November 2003
129
maintain110 and the flexibility that they have to determine the amount that they add to their
pre-paid stored value, it is not clear that redeemability attracts quite the consumer
protection premium as it does in relation to larger sums. The requirement that pre-paid
funds be redeemable has required (and will require) some redesigning of ECS-related prepaid systems in circumstances in which it is far from clear that such functionality is required
by or of significant value to customers.
We believe that issues resulting from the redeemability requirement will be resolved by the
clarification of, and the provision of guidance in relation to, the appropriate scope of
“electronic money”, as generally accepted bearer funds.
In addition, we note that the potentially overly broad scope of the redeemability requirement
itself could be viewed as a source of the money laundering concerns that are the subject of
regulation (discussed below). Allowing funds to be non-redeemable could address
concerns that micropayment prepayment mechanisms could be used to launder funds.
Waiver
Article 8 of the E-money Directive permits authorities to waive application of some or all of
the provisions of the e-money Directive. The pre-conditions for such a waiver require that
electronic money liabilities not exceed €5 million in the ordinary course (or €6 million in any
circumstances). Any approach to identifying whether stored value is electronic money that
is over-inclusive (e.g., treats all pre-paid liabilities as electronic money unless and until it is
used for another purpose) virtually guarantees that the activities of ECS providers in
relation to their pre-paid customers will not meet the quantum pre-conditions for waiver.
In addition to considering the quantum of the waiver pre-condition described above, and its
appropriateness in the context of an inclusive interpretation of the electronic money
concept, the Commission might also consider whether it is appropriate to maintain the
current criteria for identifying the electronic money ‘closed user groups’that are eligible for
waiver. Such eligibility is currently restricted to physical connections or financial or
business relationships evidenced by marketing or distribution schemes. In an ECS
environment, there are other business relationships between various players in the value
chain that are at least as close as those found in relation to financial or business
relationships. The Study Team can see no prudential nor consumer protection basis for not
treating such relationships in a similar fashion.
110
We understand that the average pre-paid quantum across Europe is currently less
than €15.00.
Ovum
November 2003
130
The potentially broad application of the “electronic money”concept (to all stored value, until
it is used for a purpose that precludes the treatment of such value as e-money)
exacerbates the potential inadequacy of the quantum pre-condition for the grant of waivers.
As such, the appropriateness of the quantum would be assisted by the clarification of the
appropriate scope of the “electronic money”concept, as generally accepted bearer funds.
In addition, the scope of the eligible financial or business relationships should be
reconsidered, in light of the growing development of electronic and other ‘virtual’business
relationships.
Customer identification for money laundering purposes
The customer identification obligations that have been imposed under the national
implementation of the Money Laundering Directive in all Member States, particularly in
relation to non-face-to-face transactions, are quite onerous.
When one considers that micropayments are, inherently, extremely small value
transactions and the limitations that result from limits imposed on maximum value that can
be held on (or added to) an ECS pre-paid balance at any time operate as a further
constraint on the possible quantum of transactions. As such, it is difficult to see how ECS
pre-paid mechanisms could be used as an effective money laundering avenue. Such an
opportunity would be further constrained if pre-paid funds accepted only by the value issuer
(i.e., not third parties) were non-redeemable (currently not permitted by Article 3 of the Emoney Directive).
We believe that the current application of the Money Laundering Directive to
micropayments should be reviewed, to consider its proportionality and appropriateness.
Ovum
November 2003
6
131
Software supply for ECNS
6.1 Introduction
As part of our study we considered whether it appears that any new developments in
the supply of software to the ECNS industry will give rise to barriers to competition in
the next two to three years. Rather than look across the whole of this complex
industry, we have examined four developments which we judge will do most to
change the ECNS industry over the next three years. These are:
• the development of operating systems for mobile terminals;
• the emergence of wireless middleware;
• the evolution of Web Services; and
• the markets for web browsers for PCs and mobile terminals.
So the analysis takes a forward looking perspective in which we focus on software
developments which will affect competition in the supply of 3G mobile services.
Our findings are based on a combination of Ovum research and responses to our
discussions documents – notably from Microsoft and Symbian.
6.2 Wireless handset operating systems
We are about to see an explosion in the demand for 3G mobile terminals. This
prompts us to ask three questions about the operating systems which these terminals
will use:
• what operating system or systems will the terminals use?
• will any one supplier be able to leverage its dominant position in other markets to
dominate the supply of mobile terminal operating systems?
• if there is multiple supply of operating systems will this create interoperability
problems for applications developers and/or service providers?
What operating systems will the terminals use?
There are currently at least five mobile terminal operating systems available:
• the Palm operating system;
• Microsoft’s Pocket PC aimed at PDAs;
• Microsoft’s Smartphone OS aimed at more basic phones;
• the Symbian operating system which is supported by all the main handset
manufacturers; and
• a Korean standard operating system supported by Samsung and LG Telecom.
We are likely to see the number of operating systems reduce in the next two to three
years. Most observers do not expect the Palm operating system to be a serious
Ovum
November 2003
132
contender in the marketplace, given the limited commercial backing which it will
receive. In addition, observers expect the Microsoft Pocket PC operating system to be
aimed at specialist niches within the corporate market. This leaves three main
operating systems competing in the mainstream 3G terminal market:
• the Symbian operating system;
• Microsoft’s Smartphone operating system; and
• the Korean standard operating system.
Will any one supplier dominate?
In any market where Microsoft is one of the main suppliers it is inevitable that
questions about dominance will arise. However, none of those who responded to our
study believe that there are grounds for concern about Microsoft leveraging its
dominance in the desktop market to become dominant in the mobile operating system
market. It is possible, of course, that Microsoft’s Smartphone operating system will
eventually dominate the mobile market. However, our respondents have indicated
that they believe that any such dominance will be the result of delivering the best
product, rather than as a result of leveraging any pre-existing power in an adjacent
market.
The prospects for a lengthy and evenly balanced struggle for market share in this
marketplace look good:
• Microsoft has strong profit streams and global brand. At the same time, it sees
mobile devices as an important marketplace in which it must become a serious
player;
• Symbian is backed by the leading European terminal suppliers and is confident
that it can win a substantial share of the market. In particular, it believes that its
licensing model is superior to that of Microsoft. Unlike Microsoft, Symbian
licenses elements of its operating system to manufacturers and does not
mandate inclusion of unwanted components. This gives the manufacturer more
control over the features it offers and, for example, allows it to build low cost, low
feature terminals more easily; and
• the Korean terminal suppliers are beginning to challenge Nokia’s dominance in
the EU mobile terminal market. They also have considerable expertise in
supplying 3G type terminals to the Korean market, developed over the last two
years.
In addition, the software industry, which is inclined to avoid regulation as a matter of
principle, has learnt a lot from the recent anti-trust cases against Microsoft. It tells us
that it is likely to alert the competition authorities to any potential anti-competitive
practices by Microsoft at an early date in future.
Are interoperability problems likely to arise?
The existence of three competing mobile operating systems suppliers will inevitably
raise the costs of applications developers, who will need to port their software on to
three platforms rather than on to a single dominant operating system. However, we
Ovum
November 2003
133
understand that these costs are modest and that the benefits generated by competing
operating systems will outweigh them substantially.
Based on the analysis set out above we conclude that:
• the competitive rivalry between mobile operating systems suppliers will generate
significant benefits;
• the interoperability problems which a multiplicity of operating systems creates are
minor; and
• there is no need for any regulatory intervention in the supply of these products.
6.3 Wireless middleware
Wireless middleware software is designed to hide the complexity of delivering
services to different types of mobile terminals by translating between network
protocols, application protocols, user interface standards, security implementations
and so on. Accordingly, wireless middleware should help operators, service providers
and enterprises to deliver sophisticated data services cost-effectively. However, the
development of such middleware could have a significant effect on the balance of
power within the 3G services industry.
Mobile network operators and terminal suppliers are currently fighting for control over
the value added services which will use mobile networks – especially 3G mobile
networks. Microsoft, Nokia and Qualcomm have promoted their middleware platforms
heavily in recent years. In particular, they have launched major developer
programmes and seeded multiple content portals either directly or via partners. For
example, Club Nokia – Nokia’s end-user portal – has long worried mobile operators
by allowing users to download free content from the Internet to their mobile terminals
via a PC, thereby bypassing their mobile operator’s network. As fixed and wireless
Internet standards converge, it becomes easier for mobile terminal suppliers to
compete on their own terms, just by offering web content. A highly intelligent terminal
capable of full access to the Internet only requires an Internet gateway at the operator
end. In this case, the end-user is able to by-pass any control that the operator might
want to exert and can, for example, use competing web-based messaging services.
This is the same situation as in the fixed network services market.
Middleware shifts the balance of power towards the mobile operator by allowing it to
deliver multiple services to a wide range of devices. This, in turn, offers content
owners the ability to reach a large user population. Wireless middleware also
provides third party applications with managed access to the services and enablers
offered by the overall service delivery platform. These may include elements such as
messaging and alerting, billing and charging, location and personalisation. A mobile
operator can use wireless middleware to provide content owners with valuable
infrastructure services, such as content adaptation and synchronisation. Content
owners can also add value to their offerings by “wrapping” them with other enablers.
All of this makes it more likely that they will want a direct interface with the operator,
Ovum
November 2003
134
rather than to offer standalone web content, accessible via a bit transport mechanism
provided by the mobile operator.
Those who responded to our study on this subject agree with this analysis. However,
none of them believe that there are any competition problems which warrant
regulatory attention. Even if middleware eventually enables mobile operators to
provide the only attractive mechanism for value added service providers and content
providers, there is still a perception of strong competition between these operators,
which should ensure that market failures do not occur.
We conclude that the development of middleware does not raise serious barriers to
competition and does not require regulatory intervention.
6.4 Web services
Definition
Web Services (“WS”) are essentially a mechanism for integrating software systems.
They are:
• a packaging technology which is both simple and universally agreed;
• accessible over the Internet without needing technology tied to any vendor’s
platform; and
• a technology which allows applications and systems to be connected together,
independently of programming languages, hardware platforms, execution models
and program locations. The software which sits “behind” a WS could be an
Enterprise Java application or component; a COBOL transaction; or a Visual
Basic program.
What makes WSs technology exciting is that it can add value to many existing
integration scenarios. In other words, it acts as a “universal glue”. As such, the
potential economic benefits from WS are very substantial. If an enterprise can use
the same type of glue for the bulk of its integration requirements, then internal
applications, trading partner applications, online services and rented applications can
all be brought together to serve the needs of the business better.
So while wireless middleware enables inter-working between different mobile
terminals on the one-hand and different mobile network services on the other, web
services are much more general mechanisms for integrating a wide range of
software systems and allowing them to inter-operate over any kind of
telecommunications networks.
Current position
WSs are already supplied by major players such as Microsoft (with .NET), IBM, Sun
and HP as well as a large number of small start-up companies. They are a rare
example of a development where there is broad industry support for all the major
players for:
Ovum
November 2003
135
• the core WSs standards – SOAP, WSDL and UDDI; and
• the formation of the WSs Interoperability Organisation (WS-I).
However, there is still a long way to go before WS can deliver its potentially
substantial benefits. There are two main dangers:
• we will see fragmentation in the implementation of the standards; and
• the infrastructure requirements will prove too complex.
At the moment, however, the development of WS is in the best commercial interests
of all the major players.
Likely impact on competition
There are good prospects that WSs will substantially increase competition in the eservices and software industries. There are three main competition effects:
• software infrastructure product vendors will compete with service providers
(through provision of hosted WS infrastructure services) and with application
vendors (through provision of business process management and WS software);
• business application software vendors will compete with service providers
(through provision of application services) and infrastructure vendors (through
provision of business process management and WS technologies); and
• on-line service providers will compete with systems integrators (through esourcing initiatives); with application vendors (through application service
provision); and with software infrastructure product vendors (through provision of
hosted WS infrastructure).
Conclusion
Those who responded to our study on this point agree with the analysis set out
above.
We therefore conclude that:
• Web Services have the potential both to increase competition in the ECNS and
adjacent markets and to deliver substantial benefits to end-users; and
• regulatory intervention is not justified - although Member State governments
could, through their own procurement policies, play an important demand side
role in stimulating the growth of Web Services.
6.5 Browsers for mobile devices
Description
An Internet Browser interprets HTML, the programming language of the Internet, into
the words and graphics that the end-user sees when viewing a web page. The most
common browser is Microsoft's Internet Explorer, which controls approximately 80%
Ovum
November 2003
136
of the market. Netscape's Navigator is a distant second. Together, browsers and
portals111 define the way in which on-line services are accessed and delivered to endusers.
Microsoft’s domination of desktop browsers
Microsoft has used its dominant position in the desktop systems market to become
dominant in the supply of desktop browsers. Browsers for fixed Internet navigation
first emerged in the mid 1990s. Netscape (formerly Mosaic) developed its Navigator
browser in 1994, and a year later it commanded an 80% market share. However,
with the launch of Windows 95, and a web browser of its own (Internet Explorer, or
“IE”) in August 1995, Microsoft began to challenge Netscape. For a while, IE played
catch-up to Netscape, as the latter continually introduced new features. However, IE
had one major advantage. Unlike Netscape, IE was free of charge and bundled with
Windows 95. Slowly but inexorably, IE gained market share. Across several
generations of browsers, IE caught up technologically with Netscape. By late 1999,
the market share situation had reversed and Microsoft had 80% of the browser
market. Eventually, the US Department of Justice forced Microsoft to break apart its
bundled software and make its APIs more transparent. However, the damage was
done by that point. Microsoft currently dominates the market for PC web browsers
and will do so for the foreseeable future. AOL acquired Netscape in 1999.
Will Microsoft dominate browsers for mobile terminals?
Will Microsoft go on to dominate the market for wireless browsers? Does it matter if it
does so?
There are currently four major types of wireless browser implementation in circulation:
• pure WAP implementations, which implement the WAP Forum's protocol stack
and interpret content specified using wireless mark-up language (WML);
• Openwave's platform. Phone.com's UP.Browser is the most deployed wireless
browser in North America;
• i-Mode. This service, deployed by NTT DoCoMo in Japan, is probably the most
widely talked-about subject in the wireless Internet space. Its architecture is
similar to that of WAP, with which is competes; and
• Microsoft's Mobile Explorer. This is a suite of web-access technologies based
around a wireless browser, and specifically designed for deployment on SIM
smartcards in feature phones and smartphones.
These different browsers do not interoperate well. They take a “lowest common
denominator” approach to ensuring interoperability across the many different devices
which might use a mobile network. In practice, this means that wireless web
111
A Web site that is commonly used as a gateway to other Web sites.
Ovum
November 2003
137
applications and content lack richness, with crude user interfaces and a general lack
of interactivity.
Such low level interoperability does not currently matter much, given the limited
processing power, storage and memory which are available in today’s mobile
terminals. Today’s microbrowsers do not carry out any local processing of data
beyond display, because there is not enough storage space, processing power and
memory to use the extensive XML libraries that are required. However, as
capabilities slowly increase, more complex web service should become possible
which involve collaboration between both network components and device
components to create applications that run across multiple device platforms.
Realistically, consumer applications based on WSs will not become a reality for mass
market mobile devices within the time horizon of this study. The required bandwidth
will not reach the necessary speeds for three to five years and, with storage space
and processing power at a premium, device vendors are more likely to earmark
resources for their own purposes.
The position is rather different in the corporate market, where extensive use of Web
Services within the next three years is a definite possibility, especially where more
powerful computing devices are used in conjunction with a wireless LAN providing
higher bandwidth. Indeed, Microsoft is to release a version of its .NET framework for
Web Services that is aimed at mobile devices running Microsoft operating systems112.
Microsoft’s wireless device operating systems are almost unique in that they are
already capable of rendering full WAP/XML on the device.
112
The .NET Compact Framework.
Ovum
November 2003
138
Based on this analysis and on the views of those interviewed for our study, we
conclude that:
• the development of high functionality web browsers for mobile device is still at an
early stage;
• there is no clear evidence that any one of the four browsers listed above will
become a de facto standard (except those used on computer devices which use
WLANs. Here we expect Microsoft to dominate through leverage of its power in
the corporate market);
• Microsoft does not have dominance in the mobile operating systems market (or
any other segment of the mobile market) in the same way that the US Courts
have found that it does in the desktop market. Accordingly, it is not in a position to
dominate the mobile browser market through leverage of dominance in an
adjacent market;
• it is unclear whether the emergence of a de facto standard wireless browser,
rather than interoperability between competing browsers is best from a public
welfare perspective;
• it should be possible to limit the opportunities which any dominant wireless
browser supplier has to leverage dominance from this market into other, adjacent,
markets (rather as the US Department of Justice has limited Microsoft in the PC
browser market); and
• there do not appear to be any market failures that would warrant the intervention
by Governments or regulatory authorities in relation to the development of
wireless browsers in the near future.
Ovum
November 2003
7
139
Separation of the incumbent
7.1 Options for separation
Market power within the telecommunications service industry in the EU remains firmly
concentrated in the hands of those organisations which formerly provided service on
a monopoly basis. For example:
• in all but two Member States, the leading mobile operator is a subsidiary of the
fixed incumbent
• in virtually all Member States, the fixed incumbent still supplies more than 90% of
fixed narrowband access lines
• in a number of Member States, the fixed incumbent owns or has a substantial
interest in the largest CATV network in the country.
Accordingly, it is necessary to consider whether this concentration of market power
represents a barrier to competition and, if so, what might be done to remove it. There
are two possible remedies:
• structural separation, also known as legal separation, in which the incumbent’s
businesses are split and run as separate entities with separately audited
accounts, but common ownership
• separation of ownership, also referred to as divestiture, in which the incumbent’s
businesses are split and operated under separate ownership.
There are four main practical applications of these two remedies:
• separation of the fixed incumbent’s business113 from the rest of the incumbent’s fixed
operations (to create Loopco and Netco)
• separation of the fixed incumbent into Netco (network activities) and Servco (retail
activities)
• separation of the fixed incumbent’s CATV businesses from its mainstream
telecommunications business
• separation of the fixed incumbent’s mobile subsidiary.
We analyse each of these possibilities below, drawing on the findings of the market
analysis of Chapter 2 (where we analyse barriers to competition in the supply of fixed
network services) and Chapter 3 (which looks at barriers to competition in the supply
of mobile services). We have deliberately looked at the options for separation of the
incumbent in a separate chapter because we believe that analysis of this problem
benefits from a consistent and comparative approach.
113
The network from the customer premises to the main distribution frame.
Ovum
November 2003
140
7.2 Divestiture of Loopco
Rivals to the fixed incumbent in countries such as Italy, Portugal and the UK have all
suggested that the fixed incumbent’s access network should be structurally separated from
the rest of its core business and run under separate ownership as Loopco. Proponents of
such separation argue that it removes barriers to local loop unbundling since, following
divestiture, Loopco no longer has an incentive to discriminate between customers.
However, there are strong economic arguments against such separation. In particular:
• separation increases the cost of contracting and bargaining between the separated
entities. These costs are measured both in monetary terms and in terms of the speed
at which the separated organisations can respond to changing market and
technological conditions; and
• separation means that it is difficult to co-ordinate investment. This is especially
important as the fixed network industry starts to migrate to next generation networks.
Substantial co-ordination of investment will be required between Loopco and Netco.
Opponents of such separation also argue that it will not solve the problem of low take up of
unbundled local loops. As we discuss in Chapter 2, the main reason for the lack of local
loop unbundling is the large fixed costs of backhaul and collocation.
Similarly, we cannot see a good case for legal separation. Such separation may
improve the transparency of the transfer price between Loopco and Netco. However,
legal separation does nothing to establish a market price for Loopco products, since
the owner is indifferent as to whether it makes profits from Loopco or Netco. At the
same time, the costs of separation are high. Many systems and procedures, which
currently operate across the Loopco/Netco boundary, would need to be reengineered.
We conclude that the case for both mandatory divestiture and legal separation is weak.
7.3 Divestiture of Netco
Several incumbents have looked at the possibility of splitting themselves into a separately
owned Netco (which provides network services) and a Servco (which provides retail
services). All of them declined to pursue the idea.
Divestiture of Netco makes regulation of the incumbent’s network services simpler because
Netco no longer has incentives to discriminate between customers. However, such
separation does not remove the need for the price regulation of network services.
More importantly for the fixed incumbent, and for public welfare, effective migration to next
generation networks requires substantial co-ordination of investment activities between:
• the retail business which will define, implement and sell next generation network
products and;
• the network business which will deliver the access and infrastructure required by these
products.
Ovum
November 2003
141
Such co-ordination is very difficult to achieve if the two businesses are separately owned
entities. These arguments have led incumbents to organise internally using a network/retail
split but to keep these operating divisions in common ownership.
The case for legal separation is somewhat stronger here, because the cost of
separation is lower. The operational split between Netco and Servco is much clearer
than that between the Loopco and Netco (see Section 7.2). Deciding whether to
require legal separation essentially requires that a view be taken as to whether the
additional transparency offered by legal separation is worth the effort when compared
with accounting separation.
We conclude that the case for mandatory divestiture of Netco is weak and the case for
legal separation marginal.
7.4 Divestiture of CATV businesses
The case for divestiture
The theoretical case for requiring fixed incumbents to divest themselves of their CATV
businesses is much stronger than the case for divestiture of Loopco or Netco. There are six
main arguments:
• these CATV businesses are based on network infrastructure which is separate from the
incumbent’s main telecommunications network and are usually run as separate
businesses. As a result, the economic objections to Loopco and Netco divestiture in
terms of increased contracting and co-ordinating costs do not apply, or apply in a much
weaker form
• most Member States’CATV networks cover a substantial proportion of the population
and, accordingly, have the potential to provide powerful infrastructure competition to
the fixed incumbent. See Figure 7.1
• even in countries where they are in financial difficulties it is clear that, when sunk costs
are excluded, CATV operators can generate EBITDA margins which are comparable
with those of fixed incumbents. In the UK, for example, NTL and Telewest both operate
at EBITDA margins in excess of 20% and project margins comparable with those
generated by incumbents within two to three years
• the development of VoIP technology should make the CATV operators’triple play
offering of TV services, fast Internet access and voice telephony increasingly
114
viable. Outside the UK and Spain, CATV operators offer broadband Internet
access and voice telephony over the same co-axial cable. However, acquisition of
these services requires two modems – a relatively cheap modem (less than €100)
for broadband and a relatively expensive modem (around €400) for voice. ETSI
has recently approved a standard for a packet modem which provides voice
telephony and broadband access in one. As a result, CATV operators expect to
114
Where Siamese pair technology is used for the final drop to the customer.
Ovum
November 2003
142
begin using the new equipment in commercial services within 12 to 18 months,
reducing by 50% to 80% the cost of the necessary end user hardware.
• fixed incumbents which own CATV networks do not have incentives to upgrade these
networks to deliver triple play offerings. Such upgrades would require scarce capital
expenditure to be used to create a major rival to their own core business. This effect is
illustrated in Figure 7.2, where the take-up of broadband services by subscribers to
CATV networks which are independently owned and those which are owned by the
fixed incumbent is compared
• divestiture of the fixed incumbent’s CATV assets could lead to stronger TV
platform competition. The next generation of DSL products, VDSL, will enable
incumbent operators to offer high quality TV services. This application will
probably be one of the driving forces for VDSL deployment. Incumbents which
own CATV networks have significantly weaker incentives to deploy VDSL
services, which potentially cannibalise their CATV business revenues, than
incumbents which do not own CATV networks.
Figure 7.1 The coverage and size of CATV networks in the EU
Member state
Austria
Belgium
Denmark
Finland
France
Germany
Greece
Ireland
Italy
Netherlands
Portugal
Spain
Sweden
UK
% homes passed
by CATV network
53%
100%
71%
59%
35%
83%
na
78%
1%
94%
60%
29%
65%
50%
% growth in Subscribers
last 4 years
2001 (m)
17%
3%
8%
14%
58%
21%
na
42%
na
7%
190%
1200%
9%
80%
1.2
3.8
1.1
1.0
3.4
21.8
na
0.6
0.1
6.2
1.1
0.5
2.1
3.6
Figure 7.2 Use of CATV networks for broadband
Measure for all OECD countries
Independent
CATV operators
CATV operators
owned by
incumbents
CATV subscribers as % of households passed – 2001
56%
50%
Cable modems as % of households passed – 2002
11%
3%
Ovum
November 2003
143
The current position
Article 8 of the Competition Directive115 requires incumbent operators to implement
legal separation, but not divestiture, of their CATV businesses when the incumbent:
• is controlled by the member state or benefits from special rights and
• is dominant in a substantial part of the market for public telecommunications network
and voice telephony and
• operates a CATV network which has been established under special or exclusive rights
in the same geographical area.
Fixed incumbents continue to own CATV networks in Denmark, Finland, France,
Luxembourg and Portugal and until very recently owned them in Germany and
Sweden as well. The same situation exists in some of the accession countries. For
example, the fixed incumbent, Matav, operates CATV networks which supply around
20% of the 1.6 million CATV subscribers in Hungary.
In Denmark, TDC, the fixed incumbent, owns the main CATV network which passes
1.3 million of Denmark’s 2.6 million households and provides service to 0.9 million of
them. There is very little overlap with the network of its main rival, TeliaSonera, which
passes 0.6 million households and serves 0.2 million of them. Neither operator offers
voice telephony over their network. Both offer cable modems. It is interesting to note
that while 13% of TeliaSonera subscribers take this service, only 3% of TDC’s
subscribers do the same.
In Portugal, PT Multimedia, a subsidiary of Portugal Telecom, owns the biggest
CATV network which now passes 2.3 million of Portugal’s 5.6 million households. Its
main rival is Cabovisao whose network passes 0.7 million homes. PT Multimedia and
Cabovisao compete in the same areas for customers. The number of their
subscribers has grown rapidly – from 0.4 million in 1997 to 1.1 million in 2001.
CATV networks in Sweden pass 65% of the 5 million households there. The market
is dominated by ComHem. This company was, until April 2003, owned by Telia.
However, the European Commission required Telia to divest itself of ComHem as a
condition of its merger with Sonera. ComHem serves 1.4 million of the 2.7 million
households passed. Its nearest rival is UPC. Take-up of cable modem service on the
UPC network is at 15%, while take-up on the ComHem network is at less than 3%.
In France, cable TV networks pass 9 million of the country’s 25 million households.
The infrastructure is largely owned by France Telecom through its subsidiary France
116
Telecom Cable , providing the cable network over which the NCNumericable
provides services, and through its minority interest in Noos117. The ART recently put
115
Which is essentially a rewording of key elements of the 1995 Cable Directive.
116
Which provides direct retail services.
117
Which uses a mix of its own network and that of France Telecom to provide
CATV services.
Ovum
November 2003
144
forward proposals for future development of the CATV sector in France. There are
three main scenarios:
• consolidation of current plans with groupings around a new entrant who could
provide fresh capital
• separation of infrastructure and service with a single infrastructure operator
providing unified access to all 9 million households for the various CATV service
providers
• the status quo.
In Germany, CATV network penetration is high. CATV networks pass 29.5 million of
the 36 million households and there are 22 million subscribers. Historically, Deutsche
Telekom dominated the industry. Until 2000, it owned 100% of the Level 3
components of these networks118 and held a minority stake in some Level 4
networks119. In 2000, under Government and EU pressure, it sold off its CATV
interests in two of the nine regions, in 2001 it sold its interest in a third region, and in
early 2003 it completed the sale of its remaining CATV interests. In comparison with
other countries where CATV networks are owned by independent operators, take-up
of voice telephony and cable modems is insignificant in Germany. Figure 7.3
illustrates. In our view, there are two main reasons for this. First, there was no
incentive for Deutsche Telecom to invest in upgrading the Level 3 networks to enable
triple play offerings. The short period since full divestiture is insufficient to assess
whether the new owners will make this investment. Secondly, co-ordination of
investment and operations is required between the Level 3 and Level 4 operators.
Where these are in separate ownership, which remains the case in many areas, there
is a further barrier to the development of effective competition to Deutsche Telekom.
The prospects for the newly independently CATV networks in Germany providing
stronger competition to Deutsche Telecom are, at least in the short term, relatively
poor:
• many of the Level 4 owners are small local organisation which seek to preserve
the local identity of their service offerings
• co-ordination of investment and operations between Level 3 and Level 4
operators remains a major problem
• the national competition authority has consistently opposed unifying the industry
around a single player120
• many observers believe that the current owner of the bulk of the assets, a
financial institution, will sell the assets within the next two years, without making
118
Which provide connection from the cable head-end to the local distribution
network.
119
Which provide the final access to the subscriber.
120
For example in its attitude towards the attempted purchase of Deutsche Telekom
assets by Liberty
Ovum
November 2003
145
the substantial investments needed to upgrade the networks so that they can
provide triple play offerings.
Figure 7.3 Take up of voice telephony and cable modem services
Country
Basic CATV
subscribers (m)
2001
Voice telephony subs
as % of CATV subs
2001
Cable modem subs
as % of CATV
network 2002
Germany
21.8
0.1%
0.2%
UK
3.6
150%
22%
USA
69
2%
16%
Conclusion
Overall, the public welfare case for requiring incumbents to divest themselves of their
CATV network businesses is a strong one and one which commands general support
amongst those we talked to during this study. Such separation can be achieved at
relatively low cost, to create an alternative infrastructure which can compete strongly
with the incumbent in the supply of broadband services and voice telephony. In
addition, as Figure 7.4 shows, the countries in which fixed incumbents own
substantial CATV network assets are all countries where CATV networks serve a
substantial proportion of the population.
Figure 7.4 CATV coverage in countries with substantial incumbent ownership
Country
% of households
passed by CATV
networks
% of CATV
networks owned by
incumbents (1)
Denmark
71%
>50%
Finland
59%
30%
France
35%
>50%
Germany
83%
>70% (2)
Portugal
60%
>70%
Sweden
65%
40% (2)
Average
60%
>50%-
Average – other member states
36%
0%
(1) As measured by number of CATV subscribers
(2) Until 3/2003
Ovum
November 2003
146
We therefore recommend that:
• the EU and individual member states consider what steps they can take to ensure
that the fixed incumbents divest themselves of their CATV businesses
• the European Commission clarify Article 23 of the universal services directive.
Article 23 requires that:
“undertakings providing telephony services at fixed locations take all reasonable
steps to ensure uninterrupted access to emergency services”.
CATV operators wishing to offer triple play services are concerned that this clause
could be interpreted as a requirement to supply voice telephony over its triple play
offering during a power cut. Such an interpretation would substantially reduce the
competitive effectiveness of CATV operators, since they would be unable to fulfil this
obligation.
If Member States are able to require the proposed divestiture, they will need to
consider:
• whether divestiture should extend to the incumbent’s CATV networks in areas
where it already competes with one or more CATV networks operators. In such
cases, it is possible that divestiture might reduce competition to the incumbent’s
mainstream offering
• whether divestiture should apply in countries, like Finland, where both the CATV
and the fixed telecommunications industry are highly fragmented.
We suggest that such issues are dealt with on a case-by-case basis.
7.5 Divestiture of mobile subsidiaries
Figure 7.5 shows that, with the exception of Ireland and the UK the mobile subsidiary
of the fixed incumbent is the largest mobile operator in every Member State. Some
incumbents, like BT in the UK and Eircom in Ireland, have voluntarily divested
themselves of their mobile subsidiaries. But is there a case for mandating such
divestiture?
There are five main arguments in favour of such divestiture:
• divestiture would lead to greater competition in retail mobile service markets. Following
divestiture, there are strong incentives for the fixed incumbent to act as an MVNO or
independent service provider on one of the mobile operator’s networks. For example,
BT has taken this course in the UK
• divestiture would increase the prospects for infrastructure competition between
fixed and mobile operators for voice traffic. Currently, there is relatively little
competition between fixed and mobile networks for voice traffic in the EU.
According to one study,121 only about 1% of such traffic transfers from the fixed to
121
Mobile and Internet substitution; Enders Analysis, 11/02.
Ovum
November 2003
147
the mobile network each year. In contrast, there is much stronger competition
between fixed and mobile operators in the USA, where the number of fixed
network long distance minutes is falling at 6% p.a. – largely as a result of mobile
substitution. An obvious barrier to such competition in the EU is the fact that the
fixed incumbent owns one of the main mobile operators in many Member States.
It is clearly not in the interests of such players to increase competition between
fixed and mobile networks for voice traffic. In contrast, in the USA, where the
main competition is for long distance minutes, five of the six largest mobile
operators are independent of long distance carriers or are owned by ILECs122.
• divestiture would increase the level of competition in the provision of public WLAN
services. Mobile operators and the fixed incumbent are obvious suppliers of such
services.
• divestiture would increase the incentives for innovation in the development of
integrated fixed and mobile services. Without a substantial mobile subsidiary, a fixed
operator concerned about loss of traffic and revenues to mobile operators, is more
likely to launch such services in an attempt to reverse or stem this trend. BT is one of
the few fixed incumbents in the EU without a mobile subsidiary. It was the first to
announce firm plans for a service which provides, through a single subscription,
integrated fixed voice telephony, private WLAN and public WLAN service in ‘hotspots’
and mobile phone service elsewhere
• divestiture makes regulation of deals between the fixed incumbent and mobile
operators more transparent and easier to regulate. The fixed incumbent would no
longer have any incentive to discriminate between mobile operators in the supply
of service bundles or of network infrastructure e.g., for next generation and 3G
networks. In addition, mobile operators would compete on equal terms to supply
the fixed incumbent with the inputs needed for services like the BT initiative
described above.
In addition, the mobile subsidiaries of the fixed incumbents are run as separate businesses
operating on separate networks. As a result, the economic objections to Loopco and Netco
divestiture do not apply.
122
The sixth, Sprint PCS, is quoted separately on the US financial markets as a
tracker stock. Sprint PCS reports separately from the rest of Sprint on its
performance to the investor community. This then leads to changes in the price of
the tracker stock. But the shareholder must trade Sprint shares as a whole and
cannot trade Sprint PCS tracker stocks on their own.
Ovum
November 2003
148
Figure 7.5 The ownership of the mobile operators in the 15 Member States
Country
Mobile subsidiary of incumbent
Market share
biggest operator
incumbent’s mobile
subsidiary
Austria
Mobilkom
45%
45%
Belgium
Proximus
54%
54%
Denmark
TeleDanmark Mobile
44%
44%
Finland
TeliaSonera
54%
54%
France
Orange
49%
49%
Germany
T-Mobile
49%
49%
Greece
Cosmote
41%
41%
Ireland
None
56%
na
Italy
TIM
49%
49%
Luxembourg
PTT GSM
65%
65%
Netherlands
KPN
41%
41%
Portugal
TMN
48%
48%
Spain
Telefonica Moviles
54%
54%
Sweden
Telia Mobitel
47%
47%
UK
None
27%
na
Source: Mobile@Ovum
Opponents argue that the benefits of divestiture are not proportionate to the disruption
which it would cause. In particular they argue that:
• the mobile subsidiaries of the fixed incumbents cannot control the degree of
competition between fixed and mobile operators in a Member State. If one of more of
the other mobile operators in the country decides to compete aggressively to win voice
traffic from the fixed networks, the fixed incumbent’s subsidiary will need to respond.
For example, the greenfield 3G mobile operator “3”is now setting prices for voice
services in Austria, Italy and the UK which suggests that it is following such a strategy
• the impact of divestiture on competition in the public WLAN market will be limited. In
any case, this market is currently small and fragmented and future demand is still
highly uncertain. For example, Ovum projects an EU public WLAN market worth €0.5
billion Euros in 2008. This compares with a current telecommunications market worth
an annual €300 billion
• the need for divestiture to ensure that the fixed incumbent does not discriminate
between mobile operators is not proven. There are few, if any, complaints at the
moment. The move to next generation networks may create new problems. However,
at this stage any such suggestions are merely speculative
• the impact of divestiture on innovation by the fixed incumbent is equally
speculative. We will need to wait several years to see whether divestiture in the
UK produces real benefits.
Ovum
November 2003
149
Based on this analysis, we conclude that there is currently no case to require
incumbents to divest themselves of their mobile subsidiaries.
The case for or against legal separation (as opposed to divestiture) is less clear-cut.
Some incumbents, such as France Telecom123 and Telecom Italia, already run their
mobile businesses as separate entities. Others do not. There is a case to require
legal separation on the grounds that it increases the transparency of deals done
between the fixed incumbent and its mobile subsidiary. This case may strengthen if
demand for integrated fixed and mobile service packages grows rapidly. However,
many argue that such problems can be dealt with through accounting separation as
easily as through legal separation.
We recommend that NRAs monitor closely the extent to which fixed incumbents and
their mobile subsidiaries favour each other to the exclusion of mobile rivals and
consider legal separation as a remedy in response to such behaviour.
123
In France Telecom’s case, the ownership is also slightly different. 17% of Orange
is owned by private shareholders who trade shares on the stock market.
Ovum
November 2003
8
150
Measures for effective competition
8.1 Introduction
In trying to enable the growth in public welfare benefits which competition in the
ECNS markets can bring, NRAs are constantly faced with the problem of trying to
determine the proper relationship between measures designed to promote
infrastructure-based competition and measures designed to promote servicebased competition. Many of the respondents to our study raised this issue
unprompted. So in this chapter we:
• assess the relative merits of infrastructure-based and service-based competition
• look at the factors which restrict the viability of infrastructure-based competition
and its future development
• consider the role of service-based competition as a complement to infrastructurebased competition
• identify the factors which NRAs need to take into account to achieve the right
balance between service-based and infrastructure-based competition
• recommend 11 measures which are designed to increase effective competition in
the ECNS markets of the EU by maximising the opportunities for infrastructurebased competition to develop where it is viable.
8.2 Definitions
We do not see infrastructure and service-based competition as clearly delineated
categories to which specific forms of competition can be allocated without ambiguity.
Instead, we use these terms in our analysis to label the opposite ends of a range of
possible competitive positions, as Figure 8.1 illustrates. It shows some of the
different ways in which service providers might provision fixed narrowband services in
competition with the incumbent, ordered by reference to whether they are closer to
pure service-based or pure infrastructure-based competition.
Ovum
November 2003
151
Figure 8.1 The possible ways of competing in the voice telephony market
Pure service-based
competition
Resale of in incumbents end to end services
eg UNE-P
Wholesale line rental + call origination +
carrier selection + call termination
Own long haul facilities + call origination +
carrier selection + call termination
Use of bitstream service to provide voice
telephony using VoDSL
Use of ULLs to provide voice telephony using
VoDSL
Pure infrastructurebased competition
AltNet (eg CATV operator) offering voice
telephony as part of triple play package
8.3 Infrastructure-based competition is better
Where it is viable, infrastructure-based competition generates more economic
benefits than the same level of service-based competition. Pure service-based
competition puts pressure on the incumbent in terms of retail efficiency, customer
service innovation and price levels. Infrastructure competition does the same, but
creates additional pressures on the incumbent to innovate in network services, to
differentiate in terms of products and pricing structures and to improve overall cost
efficiency and quality of service. In turn, this additional competition generates
increased economic benefits.
This theoretical argument commands general support amongst most regulators,
operators and analysts. There is also empirical evidence to support it. For example:
• the introduction of infrastructure-based competition in the supply of mobile
services in the late 1980s and early 1990s led to much faster take-up of mobile
services, creating clear economic benefits across the EU
• broadband take up is generally higher in those member states where the CATV
operator offers strong competition to the incumbent, as Figure 8.2 illustrates124.
This figure suggests that incumbent’s roll-out mass broadband services more
rapidly in countries where they face strong competition.
124
We have excluded Ireland from this analysis given the very early stages of
development of the broadband markets there. The % market non-incumbent includes
broadband services based on ULLs.
Ovum
November 2003
152
Figure 8.2 Infrastructure-based competition and broadband take up
Broadband services per 100 population
10.0
B
9.0
S
DK
NL
8.0
US
7.0
Jap
A
6.0
SF
5.0
D
E
4.0
F
2.0
P
UK
3.0
I
L
1.0
0.0
0%
10%
20%
30%
40%
50%
60%
70%
80%
% market non incumbent
This argument does not imply that service based competition should not co-exist with
infrastructure based competition. It is, for example, clear that competition between
service providers, as well as infrastructure based competition, has played an
important role in the development of mobile markets in many member states.
Infrastructure-based competition generates dynamic economic benefits. The
existence of infrastructure-based rivals to the incumbent (which we refer to from now
on as AltNets) gives incumbents incentives to roll out new services, to innovate in
products and prices and to invest in new technologies to become more cost efficient
at the network level. From a public welfare perspective, such incentives are
extremely important in an industry like telecommunications which is characterised by
rapid technological improvements. The incumbent does not have the same
incentives when it faces only service-based competition.
It is important that NRAs take account of these dynamic effects when making
regulatory decisions and especially when taking pricing decisions which send “build or
buy” signals to entrants. Most NRAs set prices at levels which maximise static
allocative efficiency125 Raising regulated prices from this level would lead to:
• weaker incentives for cost efficiency by the incumbent, and/or supra-normal
profits if it were to increase it’s efficiency
125
eg by setting prices to recover the long-run incremental costs of an efficient
operator with an appropriate mark-ups for common costs.
Ovum
November 2003
153
• greater investment by entrants, who build infrastructure which they would have
rented if prices had been lower126
But raising prices could also lead to more infrastructure-based competition and the
dynamics benefits which this brings. In circumstances where infrastructure-based
entry is unlikely pricing based on static allocative efficiency is the correct procedure.
But in cases where infrastructure-based entry is possible and may be even more
where preservation of infrastructure-based competition is an issue, then it is important
to consider the dynamic as well as static allocative effects.
Infrastructure based competition also requires less regulatory intervention than
service based competition. Pure infrastructure based competition may require
regulation of interconnect conditions of network operators with SMP and of the call
termination charges of individual networks. But it does not require NRAs to set the
prices at which network operators should supply service providers with inputs. This is
usually the case for service based competition. So with infrastructure based
competition there is less scope for regulatory error and for the distortions to
investment patterns which such errors bring.
8.4 Limitations on infrastructure-based competition
Unfortunately, infrastructure-based competition is not always viable, as experience in
the EU over the last ten years has demonstrated. It would seem that the cost
structures of cellular mobile networks, which require relatively modest investment in
radio access networks to create a minimum coverage network, allow strong
infrastructure competition in the supply of mobile services. In contrast, infrastructurebased competition in the supply of fixed network services is much more limited.
Why? There are two main possible explanations.
First, there are substantial economies of scale and scope in providing the fixed
access network - when compared with the radio access networks of the mobile
operators. These create a natural monopoly in the supply of certain access network
components and services. The boundaries of this natural monopoly are uncertain:
• the requirement on the fixed incumbent to rent its local loops at cost oriented
prices was a measure based on the assumption that there was a natural
monopoly in the “final mile”of the fixed network. The market reaction to local
loop unbundling, as described in Chapter 2, suggests that this natural monopoly
127
may extend further into the fixed network
126
There are of course limits to how much more the entrant would invest as the price
is increased. These limits are determined largely by the extent to which the entrant
believes it can make money in the market through its own infrastructure investment.
127
ie the high cost of backhaul and collocation make the use of unbundled local
loops unattractive for most service providers
Ovum
November 2003
154
• the boundaries to any natural monopoly may well change over time, as new
technologies change the economics of fixed access network supply.
If there are natural monopolies in the provision of fixed network services regulatory
authorities need to shift the focus of their enquiries from:
“How do we get more competition?” to
“What kind of competition is possible?”
Secondly, it is possible that the investment risks for Altnets are too high.
Organisations are not willing to make the scale of investment required to reach the
point where the AltNet’s unit costs are competitive with those of the incumbent, given
that:
•
the required investment, once made, is sunk and cannot be used for other
purposes.
•
there are substantial uncertainties in relation to customer demand and the time
required to build a customer base and, accordingly, the positive cash flow
required to earn a return on the investment.
In these circumstances, measures designed to promote service-based competition
can, at least in theory, lead to infrastructure-based competition in the long-term. By
competing at the service level, entrants can build a customer base and revenues with
little investment risk and then migrate the customers to their own facilities. This is an
attractive idea, but it has potential dangers. An NRA might attempt to implement this
“stepping stone to infrastructure-based competition” concept by setting low wholesale
prices to encourage service-based competition. However, if it sets these prices too
low, it runs the danger that it will kill infrastructure investment incentives. We set out,
in Section 2.6 (Constraint 4), a discussion of the factors which an NRA needs to take
into account when setting prices for bitstream access and other new technology
wholesale products so as to avoid such dangers.
8.5 Prospects for infrastructure-based competition
Infrastructure-based competition in fixed network services exists where operators
have found ways:
•
to overcome the economy of scale advantages of the fixed incumbent. For
example, CATV operators provide infrastructure-based competition in the supply
of broadband and voice telephony services because they can exploit the
substantial economies of scope which follow the supply of TV services.
•
to exploit high profit niches which exist when the incumbent supplies services at
prices which are averaged over different customer types and geographic areas.
This is one of the reasons, but by not means the only one, for the existence of
companies which provide telecommunication services to large corporate
customers. Such firms focus on providing large bundles of services to sites
Ovum
November 2003
155
clustered together in city centres and avoid serving low spending customers in
rural or suburban areas.
The prospects for increased infrastructure-based competition to the fixed incumbent
are not especially good, because:
• the operators that specialise in providing services to corporate customers, such
as Equant and Colt, generate EBITDA margins in the 5% to 10% range, whilst
their incumbent rivals generate margins of approximately 30%. There are few
incentives for additional investment at such margins and there is little sign of
improved margins in the near term in the projections of the financial analyst who
follows these companies.
•
the biggest of the CATV operators – companies like UPC, Telewest and ntl –
have all gone through financial difficulties recently. They have undergone
substantial financial restructuring which has effectively written off much of the
debt incurred in building their network or in upgrading them for
telecommunications purposes. Now that this restructuring is complete, they
generate reasonable margins which enable them to compete vigorously with the
incumbent within the existing footprint of their networks. However, the prospects
for their expanding this footprint into new areas are limited – at least in the short
to medium term
•
it is possible that new technologies such as fixed wireless access and powerline
will lead to substantial increases in infrastructure-based competition. But
prospects here are mixed:
-
on the one hand competitive fixed wireless access for narrowband services
has failed in the EU while the unit costs of fixed wireless access for
broadband services are typically two to four times those of DSL services
-
on the other there is now considerable renewed interest in the deployment of
powerline technologies. Various companies throughout the EU have
attempted commercial deployment of powerline technology for the past 10
years. So far they have encountered technical and regulatory problems
which have prevented mass deployment. But now a second generation of
powerline communications equipment has, once more, raised interest in this
technology and there are new initiatives to remove barriers to its deployment.
Current levels of investment in alternative infrastructure are probably at an all time
low, following reaction to the “irrational exuberance” of the investment in
telecommunications in the years leading up to 2001. We can expect to see modest
increases over the next few years, but it is unlikely that we will again see investment
plans on the scale of those made in 1999 (at the time of the last EU review of
telecommunications competition).
Ovum
November 2003
156
8.6 The role of service-based competition
It makes sense for NRAs to introduce ex ante measures which enable service-based
competition at the retail level where infrastructure-based competition is not viable. For
example:
• EU-wide requirements for fixed incumbents to offer carrier pre-selection and call
origination services have boosted service-based competition in the narrowband
calls markets
• in some Member States, NRAs require the incumbent to offer wholesale line
rental so as to generate retail competition for complete narrowband service
packages
• in some Member States, NRAs require the fixed incumbent to supply DSL access
at regulated wholesale prices, so as to enable retail competition in broadband
access services between ISPs.
The first of these measures has had a considerable impact, leading to vigorous
competition in the narrowband calls market and forcing fixed incumbents to rebalance
narrowband prices to more economically efficient levels.
Measures designed to promote service-based competition tend to have a rapid and
visible effect on retail competition in telecommunications. They might also provide a
stepping stone to infrastructure-based competition, as discussed in the previous
section. As such, they are an attractive option for NRAs.
However, such measures can also act as a roadblock – both to infrastructure-based
competition and to infrastructure investment overall. For example:
• the introduction of carrier pre-selection requirements has reduced incentives for
CATV operators to invest in voice telephony services. CATV operators may
escape SMP designation. But the existence of offers from other service based
competitors based on carrier pre-selection reduces the incentive for CATV
operators to invest in voice telephony services; and
• a requirement that fixed incumbents offer bitstream services at prices based on
simple LRIC would, as we described in Section 2.6 (Constraint 4) , reduce
substantially the incentives for investment in broadband rollout.
Accordingly, NRAs face a difficult task in achieving and maintaining the right balance
between infrastructure and service-based competition. Where do they implement
measures designed to promote service-based competition and where do they adopt
measures to promote infrastructure-based competition?
Ovum
November 2003
157
On the basis of the analysis set out above, we believe that NRAs should take account
of seven main factors when answering such a question:
a) the inherent superiority of infrastructure-based competition where it is viable and
the need to take account of the dynamic economic benefits which infrastructurebased competition brings
b) the need to take account of the tendency for a natural monopoly to exist in the
supply of fixed access network components and services, even when the extent
of this natural monopoly is unclear and/or changes over time
c) the need to set regulated prices for new technology services which do not remove
the incentives for infrastructure investment by the incumbent
d) the need to consider the possibility that service-based competition will act as a
“stepping stone”to infrastructure-based competition
e) the need to avoid service-based competition measures which undermine viable
infrastructure-based competition
f)
the relatively poor prospects for investment in further infrastructure-based
competition in fixed services
g) the fact that a significant proportion of this competition is currently based on price
averaging by the incumbent, both in terms of geography and customer groups.
8.7 Maximising infrastructure-based competition
Given the factors set out at the end of the previous section, what measures can be
taken to maximise effective infrastructure-based competition? We set out below a
series of recommendations. They are grouped under three main headings:
• measures designed to increase incentives for investment in infrastructure –
whether by incumbents or AltNets
• measures designed to increase cross platform competition between existing
infrastructure-based operators
• measures designed to increase incentives for investment by Altnets.
In formulating these measures we are conscious of the need not to encourage
128
inefficient investment in infrastructure . The recent over-investment in panEuropean broadband networks is a good example of such behaviour. But it is worth
noting that this over-investment took place in an unregulated market.
128
While allowing for the dynamic benefits which infrastructure based competition
can bring as described in Section 8.3
Ovum
November 2003
158
8.8 Increasing infrastructure investment
Measure 1: consistent regulation. Regulatory authorities should ensure that they
pursue a consistent long term regulatory policy towards infrastructure-based
operators.
It is important to start by re-stating the obvious. Infrastructure-based operators need
to be certain that the framework of rules within which they compete will not change
materially over their investment horizon, so that they can invest with confidence. Such
consistency is especially important to infrastructure-based operators who must make
very substantial investments which, once made, can rarely be used for anything
except their original purpose.
Measure 2: adequate rewards for new service investment. NRAs should set the
prices of inputs which promote service-based competition at levels which preserve
incentives for infrastructure investment by incumbents and their infrastructure based
rivals.
For legacy products, prices based on forward looking long-run incremental cost and
mark-ups are appropriate. However, for new technology wholesale products, a
premium price is needed which takes account of investment in failed new projects,
technology improvements and uncertainties in demand. Constraint 4 of Section 2.6
discusses this point in more detail in relation to the pricing of bitstream access
wholesale products.
It is clearly important to implement measures designed to promote service-based
competition where infrastructure based competition is not viable. This usually means
setting regulated upper limits on access prices to prevent the incumbent from
foreclosing such service based competition. However we believe that it is vital to set
the prices associated with these measures at a level high enough to retain incentives
for investment in new infrastructure, by the incumbent and by new entrants alike.
Failing to do so could slow down the improvements in telecommunications
infrastructure which the EU needs to remain competitive in world markets. In addition,
if the regulated wholesale price is set too low the incumbent will be unwilling to supply
- whether to itself or to others and new entrants will not invest.
8.9 Stimulating cross platform competition
Measure 3: divestiture of CATV networks. The EU and individual Member States
should take what steps they can to require fixed incumbents to divest themselves of
any remaining interests in CATV network operators. They should consider, with a
view to increasing competition, whether such divestiture should extend to geographic
areas where there is more than one CATV operator currently providing service or to
countries where the fixed CATV and fixed telecommunications industry are both
highly fragmented.
Overall, the public welfare case for requiring fixed incumbents to divest themselves of
their CATV network businesses is a strong one which commands general support
Ovum
November 2003
159
amongst those to whom we talked to during this study. Such separation can be
achieved at relatively low cost to create an alternative infrastructure which can
compete strongly with the fixed incumbent in the supply of broadband services and
voice telephony. Section 7.4 provides a detailed analysis.
Measure 4: divestiture of mobile subsidiaries. NRAs and national competition
authorities in Member States where the fixed incumbent owns the leading mobile
operator, should monitor the possibility of leverage between these two parts of the
incumbent’s business closely. Where anti-competitive conduct occurs, they should
consider requiring structural separation or divestiture.
The current concentration of market power within the telecommunications industry,
more particularly where the leading fixed and mobile operators in a given Member
State are in common ownership, is a cause for concern. In addition, there are
possible market developments, especially the development of integrated fixed and
mobile services, which could increase these concerns. However, in our view, the
current case for increased separation between the incumbent’s fixed and mobile
businesses is not yet sufficiently strong to warrant immediate action.
Measure 5: equivalent regulation of fixed and mobile call termination charges.
NRAs should ensure that the call termination charges of fixed and mobile operators
are regulated on a consistent basis.
In the past EU NRAs have typically required fixed incumbents to set call termination
charges to recover cost and their fixed rivals to charge on a reciprocal basis. This has
usually meant charging the same price for the same service129. In contrast, NRAs
have, until recently, done relatively little to set mobile call termination prices at cost
oriented levels. As a result, mobile operators have had strong incentives to let call
termination charges rise (relative to retail charges), so that they can use the profits
made from call termination to subsidise prices in the competitive retail mobile
markets.
This difference in the regulation of the fixed and mobile industry means that, where
they do compete for customer spend, fixed and mobile operators do not compete on
equal terms. The mobile operator is able to subsidise its retail prices from its call
termination charges; the fixed operator cannot. According to one recent report,130 this
asymmetry in the regulation of the fixed and mobile industry has led to a transfer of
funds from the fixed to the mobile sectors of France, Germany and the UK equal to
€19 billion over the past five years.
In recent years, some NRAs have taken steps to control mobile call termination
charges. At the same time, the EU’s Recommendation on markets susceptible to
ex ante regulation proposes that the individual call termination services of both fixed
129
We consider under Measure 7 whether this is an appropriate requirement
130
“How mobile termination charges shape the dynamics of the telecommunications
sector”, by Bomsel, Cave, le Blanc and Neumann, July 2003.
Ovum
November 2003
160
and mobile operators are distinct markets in which, by implication, the sole supplier
should be regulated in a manner that is proportionate to its degree of market power.
Measure 5 simply takes this thinking to its logical conclusion. Given the likely growing
importance of competition between fixed and mobile networks, this is an important
measure to maximise effective cross platform infrastructure competition.
Measure 6: fixed and mobile competition for corporate voice traffic. NRAs or
national competition authorities should investigate whether mobile operators are
acting in an anti-competitive manner in the pricing of the services which they offer to
large corporate customers for the termination of their voice traffic on public networks.
Fixed operators which specialise in providing large corporate customers with
telecommunications services complain that they cannot compete with the offers which
mobile operators make to their customers for terminating voice traffic from corporate
networks on public mobile networks. They allege that the mobile operators are
creating a margin squeeze. The fixed operator must pay the mobile operator’s
standard call termination charge for terminating voice traffic destined for the mobile
network. At the same time, the mobile operator charges its own corporate end users a
much lower price. Fixed operators claim that this “price squeeze” is so substantial that
they find it difficult to compete for voice traffic destined for fixed terminals as well as
for voice traffic destined for mobile terminals.
We believe that, in the long term, Measure 5 should be adequate on its own to
address any such conduct131. However, Measure 5 will take a long time to implement.
In the meantime, Measure 6 provides a relatively speedy way for regulators to
ensure, providing the case against the mobile operator is proven, that a specific
exploitation of the current asymmetric regulation of fixed and mobile call termination
charges is closed to mobile operators and that competition for corporate voice traffic
between fixed and mobile operators is increased.
8.10 Increasing investment incentives for Altnets
It is clear that the barriers to infrastructure-based competition in the fixed network
services market are substantial. The entrant does not enjoy the same economy of
scope and scale as the incumbent, nor does it own existing networks or customers.
Instead it must make major sunk cost investments in network build and customer
acquisition before it can generate revenues. This substantially raises investment
risks. In these circumstances it is important to seek measure which encourage
infrastructure-based competition as far as possible without encouraging investment
which is inefficient.
Measure 7: cost-oriented call termination charges. NRAs should allow the
infrastructure-based rivals to the incumbent – whether fixed or mobile – to set call
131
Once mobile call termination services are regulated on the same terms as fixed
call termination services then the scope for the cross subsidise as described above
should largely be eliminated.
Ovum
November 2003
161
termination charges which allow them to recover the efficiently incurred costs of an
operator of their size and topology.
Currently, most NRAs require Altnets to charge the same price as the incumbent for
the equivalent call termination service. However, it is clear that there are substantial
economies of scale in the provision of these services and that the fixed incumbent
has lower unit costs than the AltNet. These differences in unit costs arise not because
the AltNet is less efficient than the incumbent but because it is smaller as a result of
its recent entry into the market. Were the rival to grow to the same size as the
incumbent, it would have similar (or even lower) unit costs. However, the prospects of
reaching such a position are limited, as we describe in Section 8.4. In the meantime,
the current control of call termination charges means that the Altnets are unable to
recover their actual costs of call termination. Given the dynamic benefits which these
Altnets generate132, it seems to us that such regulation is inappropriate. Accordingly,
we propose Measure 7.
NRAs might object to Measure 7 on four grounds:
• it removes efficiency incentives. By setting call termination charges to recover the
costs of the most efficient operator, the NRA provides the AltNet with incentives
to become more efficient. We believe that price competition in the supply of retail
services by the AltNets provides adequate incentives for efficiency, and that this
argument ignores very real economy of scale issues that AltNets cannot address,
however efficient they may be
• it is not economically efficient. If we look from the perspective of static economic
efficiency alone then this is clearly the case. Figure 8.4 illustrates. If the smaller
rival with call termination costs C2 cannot survive it should disappear, leaving
service-based competitors to the incumbent who buy from the incumbent at the
lower price C1. In this way, we create an industry with the lowest unit cost. The
problem with this argument is that it ignores the dynamic benefits which the
existence of infrastructure-based rivals to the incumbent bring. Ultimately, of
course, it is a matter of judgement as to whether these dynamic benefits outweigh
the loss of allocative efficiency illustrated in Figure 8.4.
• it is difficult to implement in practice. Setting cost-oriented call termination
charges for a substantial number of AltNets is a time consuming burden on the
NRA and on the operators concerned. Certainly, Measure 7 involves substantial
work. However, we believe that it is a burden which the AltNets would be more
than willing to bear. In addition, given the importance of infrastructure-based
competition and the relatively small number of substantial AltNets, it is an
important issue for NRA to consider and resolve. At the same time NRAs might
focus on estimating the costs of a relatively small number of AltNets and set call
termination charges for others using this reference set of operators.
• there are other factors which lead to lower unit costs for AltNets. For example:
132
Only Altnets with directly connected customers generate call termination charges.
Ovum
November 2003
162
-
Altnets employ more modern technology and practices than the incumbent.
This is undoubtedly a factor which affects unit costs. But it is not a factor to
take into account when looking at the forward looking costs of an efficient
operator133. It is in any case a factor which has diminishing effect as
incumbents modernise their networks and support systems
-
Altnets can choose to compete only in low cost areas. This is an important
factor. But it mainly affects the unit cost of access network provision. It has
only a weak effect on the unit cost of network conveyance
-
incumbent operators may suffer the disadvantage of a legacy network load
structure which is less than optimal. So the unit costs tend to be higher than
those of an entrant. Our own cost modelling experience suggests such a
factor is relatively small when compared with economy of scale factors.
We have put the argument set out above in terms of the fixed incumbent and its
infrastructure-based rivals. However, we note that the same arguments apply to the
mobile sector, where there are also substantial economy of scale effects and where
virtual saturation in subscriber demand has made it difficult for the smaller mobile
operators to increase market share substantially.
Figure 8.4 Transferring economy of scale advantages from the incumbent
Cost of
call
termination
C2
C1
Market
share
Smaller
rival
Incumbent
Measure 8: cost benefit analysis of service-based measures. NRAs should make
an explicit, but not necessarily quantitative, assessment of the costs and benefits of
any regulatory measure which is designed to enable service-based competition.
133
The normal pricing standard when setting cost oriented call termination charges
Ovum
November 2003
163
Given the propensity of service-based measures to undermine prospects for
infrastructure-based competition, and the inherent superiority of the latter, it makes
sense for NRAs to ensure that:
• there is good reason to introduce service-based measures
• any damage which these measures do to prospects for infrastructure-based
competition is more than compensated by the benefits which they bring.
In carrying out this analysis, it makes sense for NRAs to look explicitly at short-term
and long-term effects. Service-based measures can produce immediate short-term
effects, at the expense of longer-term damage to infrastructure-based competition. In
making this assessment, NRAs will need to weight the benefits of the deeper and
stronger effects of infrastructure-based competition against the wider customer choice
which service-based competition often brings. Of course, the ideal is a market which
enjoys both strong infrastructure-based and service-based competition. But this is
often not possible.
Measure 9: preservation of geographical averaging of prices. NRAs should:
• make their current policy on geographical averaging of the fixed incumbent’s retail
prices explicit
• consider the likely impact on infrastructure-based competition before allowing the
incumbent to geographically de-average prices further.
Geographic averaging of retail prices – both narrowband and broadband – by the
incumbent is common across the EU. Indeed the normal way for an incumbent to
meet its universal service obligation to provide basic voice telephony access at an
affordable price is to offer a single geographical average price for narrowband line
rental throughout the country.
It is clear from the responses to our investigation that many in the
telecommunications sector, both NRAs and operators, believe that infrastructurebased rivals to the fixed incumbent rely on geographical averaging of the incumbent’s
retail prices for their continued survival. Further de-averaging of current prices could
weaken infrastructure-based competition while giving the incumbent freedom to fully
de-average its prices could enable it to substantially reduce the level of the
infrastructure-based competition which it faces. So while geographic averaging is not
economically efficient from a static allocative perspective, it does help maintain the
dynamic benefits of infrastructure-based competition.
Put another way, geographical averaging is a way of amplifying the effects of partial
infrastructure-based competition. Incumbents must compete with infrastructure-based
rivals, provided that the niches in which the latter operate are significant.
Geographical averaging means that the benefits of such competition, especially in
terms of price, are also enjoyed by consumers beyond the geographic area in which
the infrastructure-based competition actually occurs.
In considering arguments from the fixed incumbents for geographical deaveraging of
its retail prices NRAs will need to take account of:
• the likely impact of geographical de-averaging on the level of infrastructure based
competition
Ovum
November 2003
164
• the degree to which geographical averaging amplifies the effects of
geographically limited infrastructure based competition
• the degree to which corresponding wholesale prices are geographically
deaveraged
• social and political objectives served by geographical averaging of prices
• the economic inefficiencies introduced by geographical averaging – which sends
the wrong relative price signals to urban (low cost) and rural (high cost) areas
• the impact of geographical averaging on rollout of broadband access. If retail
broadband prices are geographically averaged then no operator may want to
serve rural areas with such services. In such circumstances revenues will be
below the cost of supply134.
Measure 10: explicit entry assistance. Given the level of infrastructure-based
competition in fixed network services, each NRA should consider whether it should
provide explicit entry assistance to infrastructure-based rivals to the fixed incumbent.
Concerned about the current low levels of investment in alternative infrastructure,
several operators, both fixed incumbents and their rivals, have suggested entry
assistance measures to boost investment. These suggestions include:
• dynamic pricing to encourage investment using unbundled local loops (as set out
in Section 2.7)
• guarantees to infrastructure-based entrants that they will avoid SMP regulation for
a specified number of years. SMP status is unlikely in the years following entry.
But entrants interviewed in our study claim that such guarantees would help them
attract funding
• tax breaks for investment e.g., in broadband infrastructure. It would be difficult to
design these tax breaks without all operators, and not just entrants, benefiting. In
these circumstances tax breaks would act as an incentive to broadband
infrastructure investment which, while it would not distort competition, might
stimulate inefficient investment.
We frame Measure 10 in terms of NRAs in individual Member States taking action,
rather than in terms of EU wide action. This reflects the fact that the level of
infrastructure-based competition varies widely across the EU, especially in broadband
markets.
Measure 11: Implementing state aid rules. The relevant authorities should ensure
that the EU rules on the use of state aid to fund telecommunications investments are
implemented rigorously.
134
There are other solutions to this problem. For example governments might (and
do) subsidise broadband roll-out in high cost rural areas, as discussed under
Measure 11. Or incumbents can deploy different products – such as fixed wireless
access or satellite technology rather than DSL over copper loops - with higher, but
still geographically averaged, prices to serve rural areas.
Ovum
November 2003
165
The European Commission recently approved the use of structural and government
funds to support broadband rollout in less favoured areas. Such funding is intended to
minimise the prospects of a digital divide emerging between the different parts of the
EU. In approving such funding, the European Commission set out conditions under
which such funds can be applied.135 For example:
• funds should support programmes which are part of coherent development plans;
• funds should be targeted at areas that would “be neglected under free market
conditions”;
• funding should only be granted where there is a competitive environment and
where the incumbent’s tariffs have been fully re-balanced; and
• funding should be technology neutral.
Despite these guidelines, virtually all respondents to our study remain concerned that
state funds will lead, and already are leading, to distortions of competition.136 They
claim that the guidelines are often ignored. All support the idea of using government
funds to minimise the digital divide between urban and rural communities, as long as:
• such funds are used only in areas where market led investment will not take
place. Funding in other areas reduces incentives for alternative infrastructure
investment. Operators point to local authorities already allegedly breaching such
constraints in places like Dublin, Norwich (UK), Amsterdam137 and Stockholm;
• such funds are used in ways which do not distort competition. AltNets are
concerned that government authorities will issue requests for proposal which, by
their size and nature, preclude AltNets from making a competitive bid, effectively
handing the contract to the incumbent; and
• government authorities, in selecting successful bidders, take account of dynamic
as well as static economic efficiencies138. Current EC guidelines explicitly prohibit
such considerations (e.g., Section 5 on Tendering Processes).
135
Guidelines on Information Society and Telecommunications Infrastructure, March
2003.
136
Broadband and Telephony Services over CATV networks, OECD, August 2003
137
Where fibre access is planned using State funds.
138
i.e. they should consider the position five years from now. If they choose the
incumbent now, will they effectively be opting for a monopoly in perpetuity with all the
loss of efficiency that would entail?
Ovum
November 2003
166
Annex A Directives affecting micropayment
services
The E-money Directive
The E-money Directive has a number of stated objectives, including:
• protecting consumers and ensuring bearer confidence;
• avoiding distortion of competition between traditional credit institutions and electronic
money institutions; and
• providing legal certainty for the development of e-commerce.
What is Electronic Money?
Electronic money is defined in the E-money Directive to mean monetary value as
represented by a claim on the issuer which is: (i) stored on an electronic device; (ii) issued
on receipt of funds of an amount not less in value than the monetary value issued; and (iii)
accepted as means of payment by undertakings other than the issuer.139 Recital 3
adds the following gloss: e-money can be considered to be an electronic surrogate for
coins and banknotes, which is stored on an electronic device such as a chip card or
computer memory and which is generally intended for the purpose of effecting electronic
payments of limited amounts.
The recitals to the E-money Directive make it clear that the issuance of electronic money
does not constitute, in itself, given its specific character as an electronic surrogate for coins
and banknotes, a “deposit-taking activity” (under the Codified Banking Directive) if the
funds are immediately exchanged for electronic money. However, the receipt of funds in
exchange for electronic money which results in a credit balance left on account with the
issuing institution constitutes the receipt of “deposits” or other “repayable funds” (under the
Codified Banking Directive).
Article 3 provides that bearers of electronic money must be able to ask the issuer to
redeem it at par value (whether in cash or through transfer to an account) free of charges
other than those strictly necessary to carry out the operation. The recitals make it clear that
the redeemability requirement is designed to ensure bearer confidence. It does not imply,
in itself, that funds received in exchange for electronic money are to be regarded as
deposits or other repayable funds.
The Scope of the E-money Directive
The E-money Directive applies to “electronic money institutions” or “EMIs” (i.e., any legal
person, other than a credit institution (as defined in the Codified Banking Directive) which
139
Article 1(3)(b).
Ovum
November 2003
167
issues means of payment in the form of electronic money).140 It requires Member States to
prohibit persons or undertakings that are not credit institutions from carrying on the
business of issuing electronic money.141
EMIs that have been found by the competent authorities of the “home”Member State to
satisfy the prudential rules set out below are authorised to issue electronic money through
the EU, either by providing such services at a distance to residents of other Member States
or by establishing a branch in those Member States (or both).
The authorisation requirement implies that schemes from third countries are not entitled to
remotely issue electronic money in the EU. However, this does not preclude the use in the
EU of electronic money issued outside the EU under such schemes.
The Impact of the E-money Directive
The business activities of EMIs other than the issuing of electronic money are restricted, by
Article 1(5) to:
• the provision of closely related financial and non-financial services such as the
administering of electronic money by the performance of operational and other ancillary
functions related to its issuance, and the issuing and administering of other means of
payment but excluding the granting of any form of credit; and
• the storing of data on the electronic device on behalf of other undertakings or public
institutions.
Article 4 imposes a minimum initial capital requirement of not less than €1 million. In
addition, EMIs must have at all times own funds which are equal to or above 2% of the
higher of the current amount or the average of the preceding six months’total amount of
their financial liabilities related to outstanding electronic money.
Article 5 goes on to impose strict restrictions on investments by EMIs. They must have
investments of no less than their financial liabilities related to outstanding electronic money
in: (a) assets that attract a zero credit risk weighting and are sufficiently liquid, (b) sight
deposits held with Zone A credit institutions, or (c) debt instruments that are sufficiently
liquid and satisfy a number of other precise requirements. Type (a) and (b) investments
may not exceed 20 times the own funds of the electronic money institution.
Article 6 requires six monthly checks to ensure that EMIs are complying with their
obligations under Articles 4 and 5.
Recital 12 of the E-money Directive sets out the philosophy behind the supervisory
provisions. It refers to the need to preserve a level playing field between EMIs and other
credit institutions issuing electronic money, to ensure fair competition among a wider range
of institutions, to the benefit of bearers. This is achieved through offsetting a lighter
prudential regime with more stringent restrictions on business activities and on investments
140
Article 1(1).
141
Article 1(4).
Ovum
November 2003
168
(to ensure that their financial liabilities related to outstanding electronic money are backed
by sufficiently liquid low risk assets). It is not entirely clear how these requirements “level”
the competitive playing field.
Finally, Article 7 requires EMIs to have sound and prudent management, administrative and
accounting procedures and adequate internal control mechanisms, appropriate to the
financial and non-financial risks to which the institution is exposed.
The Possibility of Waiver
Member States may allow their authorities to waive the application of some or all of the
provisions of the E-money Directive and Directive 2000/12/EC where:
• the total business activities related to electronic money generate a total amount of
liabilities related to outstanding electronic money that normally does not exceed €5
million and never exceeds €6 million;
• the electronic money issued is accepted as a means of payment only by subsidiaries
which perform operational or other ancillary functions related to electronic money
issued or distributed by the institution or any parent or sibling of the undertaking; or
• the electronic money issued is accepted as payment only by a limited number of
undertakings, clearly distinguishable by:
-
their location in the same premises or other limited local area; or
-
their close financial or business relationship (e.g., common marketing or
distribution scheme) with the issuing institution.142
Recital 15 emphasises that the possibility of waiving requirements exists only in relation to
EMIs which operate only within the territory of the relevant Member State.
The Electronic Signatures Directive
Electronic signatures (i.e., data in electronic form which are attached to or logically
associated with other electronic data and which serve as a method of authentication)143
serve to, inter alia, facilitate data authentication. They do so through the issuance of
“certificates” (i.e., electronic attestations that like signature-verification data to a person and
confirms the identity of that person) by “certification-service-providers”.144
While Member States are not permitted to impose a ‘prior authorisation’regime for the
certification services, they are required to establish an appropriate system that allows for
supervision of certification-service-providers which are established on their particular
territory and which issue qualified certificates to the public. As such, certification-service-
142
Article 8.
143
Article 2(1) of the Electronic Signatures Directive.
144
Article 2(9).
Ovum
November 2003
169
providers must comply with the requirements of the Member State in which they are
established. Their liability to the public for the provision of such services is also subject to
national rules.
The Codified Banking Directive
The prudential regime outlined above for EMIs differs from, although it purports to be
‘calibrated on’, that which applies to other credit institutions (under the Codified Banking
Directive). As such, only references to credit institutions in the Codified Banking Directive
(except those in Title V, Chapter 2) apply to EMIs. Articles 5, 11, 13, 19, 20(7), 51 and 59
of the Codified Banking Directive expressly do not apply to EMIs.
Directive 2000/28/EC amends the Codified Banking Directive to, inter alia, include
electronic money institutions in the “credit institution” definition, in Article 1(1)(b).
The “credit institution” definition in Article 1(1) also includes undertakings whose business it
is to grant credits for its own account. Accordingly, institutions which grant credits for their
own account must comply in full with the requirements of Title II, relating to authorisation,
initial capital, management, operations and organisation, procedures and internal control
mechanisms, and will be subject to prudential supervision in accordance with Title V.
In addition, the Protocol on the Statute of the ECB and the ESCB provides for the
imposition of reserve requirements on “credit institutions”.
The Money Laundering Directive
The Money Laundering Directive applies to all credit institutions, including EMIs. It requires
Member States to ensure that money laundering is prohibited and to impose a range of
obligations on, inter alia, credit institutions. These obligations include:
• ensuring that credit institutions require supporting evidence to identify their customers;
• ensure that credit institutions take specific and adequate measures necessary to
compensate for the greater risk of money laundering that arises when transactions are
entered into with customers who have not been physically present for identification
purposes (“non-face-to-face”operations). Such measures must ensure that the
customer’s identity is established. This could entail requiring additional documentary
evidence, supplementary measures to verify or certify the documents supplied,
confirmatory certification by an institution subject to the Money Laundering Directive, or
that the first payment of the operation is carried out through an account opened in the
customer’s name with a credit institution subject to the Money Laundering Directive;145
• establishing adequate procedures of internal control and communication to forestall
and prevent operations related to money laundering; and
145
Article 3.
Ovum
November 2003
170
• taking steps to ensure that employees are aware of the provisions of the Money
Laundering Directive, including the participation of the relevant employees in special
training programmes.
The Money Laundering Directive also imposes obligations to inform relevant authorities, on
the initiative of the credit institutions, and to provide those authorities (on request) with all
information deemed necessary to combat money laundering.
Ovum
November 2003
171
Annex B Glossary
2G: Second generation. A generic term used to describe the current generation of
digital mobile services eg those using GSM or CDMA 95 technologies.
3G: Third generation. The generic term used for the next generation of mobile
communications systems. Third-generation systems will provide enhanced services,
including higher bandwidth and connectionless ‘always on’services, thus enabling a
wider range of end-user services and charging models than are available today.
ADSL: Asymmetric digital subscriber line. An xDSL technology that can transmit
downstream rates of up to 9Mbit/s, depending on distance, and upstream rates up to
0.5Mbit/s.
Altnet: An alternative network operator who competes with the fixed incumbent and
provides direct connections to its customers.
AOL: America On line, a leading content focussed ISP
API: Application programming interface. A well-defined interface presented by one
software program that allows another program (as opposed to a human user) to make
requests of that program.
ARPU: Average revenue per user.
ASP: applications service provider
ATM: Asynchronous transfer mode. A switching technology that can handle all forms
of traffic (voice, video and data) within a single network at very high speeds.
AT&T: a US based long distance operator with a significant presence in the EU
corporate markets
BiB: British Interactive Broadcasting
BPO: Business process outsourcing.
BT: British Telecom, the UK’s fixed incumbent
CAGR: Compound annual growth rate
CAS: Conditional access system usually located in a set top box attached to a
television set to provide access to programmes only to authorised viewers
CATV: Cable television
CDMA: Code division multiple access. A method of frequency re-use whereby many
radios use the same frequency, but each one has a unique code. GPS uses CDMA
techniques with Gold’s codes for their unique cross-correlation properties.
CLEC: Competitive local exchange carrier
Ovum
November 2003
172
Cobol: A legacy high level programming language which was used extensively in the
early days of computing.
COLT: a pan European Altnet
CPE: Customer premise equipment.
CUG: Closed user group.
DDE: Declarative Data Essence Group standard
DG: Directorate General
DSL: Digital subscriber line – a technology for sending multimedia transmissions
over copper-pair telephone lines.
DSLAM: DSL access multiplexor
DTD: Document type definition
DTT: Digital terrestrial television.
DVB: Digital Video Broadcasting
e.biscom: an Italian based Altnet
EBITDA: Earnings before interest, taxation, depreciation and amortisation.
EC: European Commission
ECB: European Central Bank
ECNS: Electronic communications networks and services
ECN: Electronic communications networks
ECS: Electronic communications services
ECTA: European Competitive Telecommunications Association
EDGE: Enhanced data rate for GSM extension (EDGE). An intermediate stage for
introducing high-speed data to mobile networks ahead of 3G.
EMI: Electronic money issuer
ENUM: A mechanism for translating between circuit-switched E.164 numbers and
Internet names.
EPG: Electronic programme guide
ERG: European Regulators Group
ESCB: The European System of Central Banks
ETSI: The European Telecommunications Standards Institute
Ovum
November 2003
173
FCC: The Federal Communications Commission – the US federal regulator of
telecoms and broadcasting
FTP: File transfer protocol
GPRS: General packet radio service. GSM Phase II+ provides high-speed packet
data rates up to a theoretical maximum of 170kbit/s. GPRS is an enhancement to
cellular phone networks, as it enables them to carry packet-switched data traffic.
GSM: Global system for mobile communications. The standard for digital cellular
systems used throughout Europe and in other countries around the world, operating
in several frequency bands.
IE: Internet Explorer
ILEC: Incumbent local exchange carrier
INTUG: a user group representing the interests of EU business users of
telecommunications
IP: Internet protocol. The basic set of network-layer technologies and conventions
under which data is sent, routed and received in the Internet. It is also widely applied
in private networks (known as ‘intranets’), closed-group networks (‘extranets’) and
next-generation telecoms networks.
IPR: Intellectual property rights.
IP VPN: IP virtual private network. A closed user group on a carrier’s network that
provides IP at each access. It may use infrastructure within the carrier network that is
shared by other users.
IRG: Independent Regulators Group
ISP: Internet service provider. Service provider supplying IP-based access to the
Internet either to end-user customers or on a wholesale basis to other ISPs.
iTV: interactive television
KPN: The Dutch incumbent operator
LRIC: long run incremental cost
M6: French 6th Broadcasting Channel
MDF: Main distribution frame. The frame on which the primary cables of the local
distribution network terminate. The MDF is normally the end of the copper access
network.
MHEG: Multimedia and Hypermedia Experts Group standard
MHP: Multi media home platform
Ovum
November 2003
174
MMS: Multimedia message service. A standard defined by the 3GPP for sending
messages to and from mobile phones, which can contain text, graphics, photos, audio
and video.
MNO: Mobile network operator.
MPLS: Multi protocol label switching
MPSA: Mobile Payment Service Association
MVNO: Mobile virtual network operator.
NGN: Next generation network. Usually refers to the use of a packet switched,
rather than circuit switched network to carry multi media (voice, data and image)
content
ntl: a UK based CATV operator
NRA: National Regulatory Authority. Term used by European Commission to denote
the agency responsible for regulating a country’s telecommunications sector.
NTT: The Japanese fixed incumbent
NVoD: Near video on demand
Oftel: UK NRA
OFT: Office of Fair Trading (UK)
O2: a mobile operator in Germany, Ireland and the UK
OMA: Open Mobile Alliance
OPTA: Dutch NRA
OS: Operating system. A piece of software that administers the basic operations of a
computer.
OSS: Operational support system
PC: Personal computer
PDA: Personal digital assistant. A palm-top computer that performs specific
computing tasks, such as an electronic diary, organiser, carry-along personal
database, multimedia player and the management of all telephone-related
communications. The communications will take place through the telephone or
through wireless transmission to a cellular service or desktop system.
PLC: Powerline communication
PRS: Premium rate service
PSP: Payment service provider
Ovum
November 2003
175
PSTN: Public switched telephone network. A term for the amalgam of all the publicly
available, interconnected telco networks.
PVR: Personal Video Recorders
RAM: Random access memory
ROCE: Return on capital employed
ROM: Read only memory
RTE: Radio Telefís Éireann
SDH: synchronous digital hierarchy
SIMLOCK: a method for locking a subscriber’s SIM card for use only in a particular
handset
SME: Small or medium sized enterprise
SMP: Significant market power
SMS: Short message services. The sending and receiving of short alphanumeric
messages to and from mobile handsets on a cellular mobile network.
SOAP: Simple object access protocol
SSSL: Sky Subscribers Services Limited
STB: Set top box
Symbian: a company set up to develop a mobile terminal operating system
TDC: Telecom Denmark, Denmark’s incumbent operator
TF1: Télévision Française 1
TIM: Telecom Italia Mobile
TPS: Télévision Par Satellite
TWF: TV without frontiers
UDDI: Universal description discovery and integration
ULL: Unbundled local loop. A local loop (usually copper) that is rented by its owner
(usually the incumbent) to another operator (a competitive access provider) to create
a direct customer connection.
UMTS: Universal mobile telecommunications system. The international standard for
third-generation (broadband) mobile communications.
UNE-P: Unbundled network elements – platform. A combination of WLR and local
switching which ILECs in the USA must make available at regulated prices.
UPC: a CATV operator with interests in several EU countries
Ovum
November 2003
176
URL: Universal resource locator. A human-readable naming scheme that can be
used to address any IP-based service (including not only web-based services but also
FTP and others) over the Internet.
VAS: Value-added service.
VCR: Video cassette recorder.
VDSL: very high speed DSL
VoD: Video on demand
VoDSL: Voice over DSL
VoIP: voice over IP
WACC: Weighted average cost of capital
WAP: Wireless application protocol. A widely backed standard that supports Internet
content access and call management functionality for small-screen wireless devices.
W-CDMA: Wideband CDMA. A standard for 3G networks adopted by most European
and some US and Asia-Pacific operators.
WLAN: Wireless local area network
WLR: Wholesale line rental.
WML: Wireless mark-up language, designed specifically for WAP phones.
WS: Web services
WSDL: Web service description language
XML: Extensible mark-up language, which can be described as a generic mark-up
language solution. It differs from other mark-up solutions in that it only describes the
data on a web page and not how it should be formatted, providing the flexibility for the
data to be quickly formatted for display on other devices.
Ovum
November 2003

final report V8

Transcrição

Documentos relacionados

Untitled - 4G Americas

protocols and agreement for satellite communications on mistralis

pdf,24KB

Mídia Kit

Latin America AGENDA Wednesday 13th May 2015

Partisanship and Electoral Accountability: Evidence

Digital inclusion in Latin America and the Caribbean

1 CCT Programs, the Economy, and Presidential Elections in

Effect of handwheel sizes on the turning speed.: NA

Shift to local, shift to mobile